Netdev List
 help / color / mirror / Atom feed
* [PATCH v3 net-next 05/19] ionic: Add interrupts and doorbells
From: Shannon Nelson @ 2019-07-08 19:25 UTC (permalink / raw)
  To: snelson, netdev
In-Reply-To: <20190708192532.27420-1-snelson@pensando.io>

The ionic interrupt model is based on interrupt control blocks
accessed through the PCI BAR.  Doorbell registers are used by
the driver to signal to the NIC that requests are waiting on
the message queues.  Interrupts are used by the NIC to signal
to the driver that answers are waiting on the completion queues.

Signed-off-by: Shannon Nelson <snelson@pensando.io>
---
 drivers/net/ethernet/pensando/ionic/ionic.h   |  3 +
 .../net/ethernet/pensando/ionic/ionic_bus.h   |  2 +
 .../ethernet/pensando/ionic/ionic_bus_pci.c   | 12 ++++
 .../net/ethernet/pensando/ionic/ionic_dev.c   |  6 ++
 .../net/ethernet/pensando/ionic/ionic_dev.h   | 21 ++++++
 .../net/ethernet/pensando/ionic/ionic_lif.c   | 64 +++++++++++++++++++
 .../net/ethernet/pensando/ionic/ionic_lif.h   |  4 ++
 7 files changed, 112 insertions(+)

diff --git a/drivers/net/ethernet/pensando/ionic/ionic.h b/drivers/net/ethernet/pensando/ionic/ionic.h
index afff2e074f1a..c86a08752b3b 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic.h
@@ -4,6 +4,8 @@
 #ifndef _IONIC_H_
 #define _IONIC_H_
 
+struct lif;
+
 #include "ionic_if.h"
 #include "ionic_dev.h"
 
@@ -38,6 +40,7 @@ struct ionic {
 	unsigned int nrxqs_per_lif;
 	DECLARE_BITMAP(lifbits, IONIC_LIFS_MAX);
 	unsigned int nintrs;
+	DECLARE_BITMAP(intrs, INTR_CTRL_REGS_MAX);
 };
 
 int ionic_dev_cmd_wait(struct ionic *ionic, unsigned long max_wait);
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus.h b/drivers/net/ethernet/pensando/ionic/ionic_bus.h
index 3b1e2d0ebf8f..6b29e94f81d6 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_bus.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_bus.h
@@ -9,5 +9,7 @@ int ionic_bus_alloc_irq_vectors(struct ionic *ionic, unsigned int nintrs);
 void ionic_bus_free_irq_vectors(struct ionic *ionic);
 int ionic_bus_register_driver(void);
 void ionic_bus_unregister_driver(void);
+void __iomem *ionic_bus_map_dbpage(struct ionic *ionic, int page_num);
+void ionic_bus_unmap_dbpage(struct ionic *ionic, void __iomem *page);
 
 #endif /* _IONIC_BUS_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
index 0ed724efe117..838d7c423b2c 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
@@ -87,6 +87,18 @@ static void ionic_unmap_bars(struct ionic *ionic)
 	}
 }
 
+void __iomem *ionic_bus_map_dbpage(struct ionic *ionic, int page_num)
+{
+	return pci_iomap_range(ionic->pdev,
+			       ionic->bars[IONIC_PCI_BAR_DBELL].res_index,
+			       (u64)page_num << PAGE_SHIFT, PAGE_SIZE);
+}
+
+void ionic_bus_unmap_dbpage(struct ionic *ionic, void __iomem *page)
+{
+	iounmap(page);
+}
+
 static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	struct device *dev = &pdev->dev;
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.c b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
index 59d863e45072..c0710bf200a5 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_dev.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
@@ -9,6 +9,7 @@
 #include <linux/etherdevice.h>
 #include "ionic.h"
 #include "ionic_dev.h"
+#include "ionic_lif.h"
 
 void ionic_init_devinfo(struct ionic_dev *idev)
 {
@@ -280,3 +281,8 @@ void ionic_dev_cmd_lif_reset(struct ionic_dev *idev, u16 lif_index)
 
 	ionic_dev_cmd_go(idev, &cmd);
 }
+
+int ionic_db_page_num(struct lif *lif, int pid)
+{
+	return (lif->hw_index * lif->dbid_count) + pid;
+}
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.h b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
index a8a84734b433..6d30adeab8c5 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_dev.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
@@ -132,8 +132,27 @@ struct ionic_dev {
 	struct ionic_devinfo dev_info;
 };
 
+#define INTR_INDEX_NOT_ASSIGNED		-1
+#define INTR_NAME_MAX_SZ		32
+
+struct intr {
+	char name[INTR_NAME_MAX_SZ];
+	unsigned int index;
+	unsigned int vector;
+	u64 rearm_count;
+	unsigned int cpu;
+	cpumask_t affinity_mask;
+};
+
 struct ionic;
 
+static inline void ionic_intr_init(struct ionic_dev *idev, struct intr *intr,
+				   unsigned long index)
+{
+	ionic_intr_clean(idev->intr_ctrl, index);
+	intr->index = index;
+}
+
 void ionic_init_devinfo(struct ionic_dev *idev);
 int ionic_dev_setup(struct ionic *ionic);
 void ionic_dev_teardown(struct ionic *ionic);
@@ -163,4 +182,6 @@ void ionic_dev_cmd_lif_init(struct ionic_dev *idev, u16 lif_index,
 			    dma_addr_t addr);
 void ionic_dev_cmd_lif_reset(struct ionic_dev *idev, u16 lif_index);
 
+int ionic_db_page_num(struct lif *lif, int pid);
+
 #endif /* _IONIC_DEV_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.c b/drivers/net/ethernet/pensando/ionic/ionic_lif.c
index e32ce2a3d0ce..65ced9b469c9 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_lif.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.c
@@ -12,6 +12,30 @@
 #include "ionic_lif.h"
 #include "ionic_debugfs.h"
 
+static int ionic_intr_alloc(struct lif *lif, struct intr *intr)
+{
+	struct ionic *ionic = lif->ionic;
+	int index;
+
+	index = find_first_zero_bit(ionic->intrs, ionic->nintrs);
+	if (index == ionic->nintrs) {
+		netdev_warn(lif->netdev, "%s: no intr, index=%d nintrs=%d\n",
+			    __func__, index, ionic->nintrs);
+		return -ENOSPC;
+	}
+
+	set_bit(index, ionic->intrs);
+	ionic_intr_init(&ionic->idev, intr, index);
+
+	return 0;
+}
+
+static void ionic_intr_free(struct lif *lif, int index)
+{
+	if (index != INTR_INDEX_NOT_ASSIGNED && index < lif->ionic->nintrs)
+		clear_bit(index, lif->ionic->intrs);
+}
+
 static struct lif *ionic_lif_alloc(struct ionic *ionic, unsigned int index)
 {
 	struct device *dev = ionic->dev;
@@ -96,6 +120,12 @@ static void ionic_lif_free(struct lif *lif)
 	lif->info = NULL;
 	lif->info_pa = 0;
 
+	/* unmap doorbell page */
+	ionic_bus_unmap_dbpage(lif->ionic, lif->kern_dbpage);
+	lif->kern_dbpage = NULL;
+	kfree(lif->dbid_inuse);
+	lif->dbid_inuse = NULL;
+
 	/* free netdev & lif */
 	ionic_debugfs_del_lif(lif);
 	list_del(&lif->list);
@@ -138,7 +168,9 @@ void ionic_lifs_deinit(struct ionic *ionic)
 static int ionic_lif_init(struct lif *lif)
 {
 	struct ionic_dev *idev = &lif->ionic->idev;
+	struct device *dev = lif->ionic->dev;
 	struct lif_init_comp comp;
+	int dbpage_num;
 	int err;
 
 	ionic_debugfs_add_lif(lif);
@@ -153,9 +185,41 @@ static int ionic_lif_init(struct lif *lif)
 
 	lif->hw_index = le16_to_cpu(comp.hw_index);
 
+	/* now that we have the hw_index we can figure out our doorbell page */
+	lif->dbid_count = le32_to_cpu(lif->ionic->ident.dev.ndbpgs_per_lif);
+	if (!lif->dbid_count) {
+		dev_err(dev, "No doorbell pages, aborting\n");
+		return -EINVAL;
+	}
+
+	lif->dbid_inuse = kzalloc(BITS_TO_LONGS(lif->dbid_count) * sizeof(long),
+				  GFP_KERNEL);
+	if (!lif->dbid_inuse) {
+		dev_err(dev, "Failed alloc doorbell id bitmap, aborting\n");
+		return -ENOMEM;
+	}
+
+	/* first doorbell id reserved for kernel (dbid aka pid == zero) */
+	set_bit(0, lif->dbid_inuse);
+	lif->kern_pid = 0;
+
+	dbpage_num = ionic_db_page_num(lif, lif->kern_pid);
+	lif->kern_dbpage = ionic_bus_map_dbpage(lif->ionic, dbpage_num);
+	if (!lif->kern_dbpage) {
+		dev_err(dev, "Cannot map dbpage, aborting\n");
+		err = -ENOMEM;
+		goto err_out_free_dbid;
+	}
+
 	set_bit(LIF_INITED, lif->state);
 
 	return 0;
+
+err_out_free_dbid:
+	kfree(lif->dbid_inuse);
+	lif->dbid_inuse = NULL;
+
+	return err;
 }
 
 int ionic_lifs_init(struct ionic *ionic)
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.h b/drivers/net/ethernet/pensando/ionic/ionic_lif.h
index 37f79d36744a..98d0699234e9 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_lif.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.h
@@ -23,6 +23,8 @@ struct lif {
 	bool registered;
 	unsigned int index;
 	unsigned int hw_index;
+	unsigned int kern_pid;
+	u64 __iomem *kern_dbpage;
 	unsigned int neqs;
 	unsigned int nxqs;
 
@@ -30,6 +32,8 @@ struct lif {
 	dma_addr_t info_pa;
 	u32 info_sz;
 
+	unsigned long *dbid_inuse;
+	unsigned int dbid_count;
 	struct dentry *dentry;
 	u32 flags;
 };
-- 
2.17.1


^ permalink raw reply related

* [PATCH v3 net-next 02/19] ionic: Add hardware init and device commands
From: Shannon Nelson @ 2019-07-08 19:25 UTC (permalink / raw)
  To: snelson, netdev
In-Reply-To: <20190708192532.27420-1-snelson@pensando.io>

The ionic device has a small set of PCI registers, including a
device control and data space, and a large set of message
commands.

Signed-off-by: Shannon Nelson <snelson@pensando.io>
---
 drivers/net/ethernet/pensando/ionic/Makefile  |    2 +-
 drivers/net/ethernet/pensando/ionic/ionic.h   |   20 +
 .../net/ethernet/pensando/ionic/ionic_bus.h   |    1 +
 .../ethernet/pensando/ionic/ionic_bus_pci.c   |  140 +-
 .../ethernet/pensando/ionic/ionic_debugfs.c   |   67 +
 .../ethernet/pensando/ionic/ionic_debugfs.h   |   28 +
 .../net/ethernet/pensando/ionic/ionic_dev.c   |  132 +
 .../net/ethernet/pensando/ionic/ionic_dev.h   |  144 +
 .../net/ethernet/pensando/ionic/ionic_if.h    | 2552 +++++++++++++++++
 .../net/ethernet/pensando/ionic/ionic_main.c  |  296 ++
 .../net/ethernet/pensando/ionic/ionic_regs.h  |  133 +
 11 files changed, 3512 insertions(+), 3 deletions(-)
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_debugfs.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_debugfs.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_dev.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_dev.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_if.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_regs.h

diff --git a/drivers/net/ethernet/pensando/ionic/Makefile b/drivers/net/ethernet/pensando/ionic/Makefile
index beb3faeccac1..49a8e28a5c84 100644
--- a/drivers/net/ethernet/pensando/ionic/Makefile
+++ b/drivers/net/ethernet/pensando/ionic/Makefile
@@ -3,4 +3,4 @@
 
 obj-$(CONFIG_IONIC) := ionic.o
 
-ionic-y := ionic_main.o ionic_bus_pci.o
+ionic-y := ionic_main.o ionic_bus_pci.o ionic_dev.o ionic_debugfs.o
diff --git a/drivers/net/ethernet/pensando/ionic/ionic.h b/drivers/net/ethernet/pensando/ionic/ionic.h
index 18c79adef06c..4fd0ae86556a 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic.h
@@ -4,6 +4,10 @@
 #ifndef _IONIC_H_
 #define _IONIC_H_
 
+#include "ionic_if.h"
+#include "ionic_dev.h"
+#include "ionic_debugfs.h"
+
 #define DRV_NAME		"ionic"
 #define DRV_DESCRIPTION		"Pensando Ethernet NIC Driver"
 #define DRV_VERSION		"0.11.0-k"
@@ -17,9 +21,25 @@
 #define IONIC_SUBDEV_ID_NAPLES_100_4	0x4001
 #define IONIC_SUBDEV_ID_NAPLES_100_8	0x4002
 
+#define devcmd_timeout  10
+
 struct ionic {
 	struct pci_dev *pdev;
 	struct device *dev;
+	struct ionic_dev idev;
+	struct mutex dev_cmd_lock;	/* lock for dev_cmd operations */
+	struct dentry *dentry;
+	struct ionic_dev_bar bars[IONIC_BARS_MAX];
+	unsigned int num_bars;
+	struct identity ident;
 };
 
+int ionic_dev_cmd_wait(struct ionic *ionic, unsigned long max_wait);
+int ionic_set_dma_mask(struct ionic *ionic);
+int ionic_setup(struct ionic *ionic);
+
+int ionic_identify(struct ionic *ionic);
+int ionic_init(struct ionic *ionic);
+int ionic_reset(struct ionic *ionic);
+
 #endif /* _IONIC_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus.h b/drivers/net/ethernet/pensando/ionic/ionic_bus.h
index 94ba0afc6f38..24b4c01ec03f 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_bus.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_bus.h
@@ -4,6 +4,7 @@
 #ifndef _IONIC_BUS_H_
 #define _IONIC_BUS_H_
 
+const char *ionic_bus_info(struct ionic *ionic);
 int ionic_bus_register_driver(void);
 void ionic_bus_unregister_driver(void);
 
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
index 3fc3479c85f1..7eb23f0e87ac 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
@@ -17,27 +17,163 @@ static const struct pci_device_id ionic_id_table[] = {
 };
 MODULE_DEVICE_TABLE(pci, ionic_id_table);
 
+const char *ionic_bus_info(struct ionic *ionic)
+{
+	return pci_name(ionic->pdev);
+}
+
+static int ionic_map_bars(struct ionic *ionic)
+{
+	struct pci_dev *pdev = ionic->pdev;
+	struct device *dev = ionic->dev;
+	struct ionic_dev_bar *bars;
+	unsigned int i, j;
+
+	bars = ionic->bars;
+	ionic->num_bars = 0;
+
+	for (i = 0, j = 0; i < IONIC_BARS_MAX; i++) {
+		if (!(pci_resource_flags(pdev, i) & IORESOURCE_MEM))
+			continue;
+		bars[j].len = pci_resource_len(pdev, i);
+
+		/* only map the whole bar 0 */
+		if (j > 0) {
+			bars[j].vaddr = NULL;
+		} else {
+			bars[j].vaddr = pci_iomap(pdev, i, bars[j].len);
+			if (!bars[j].vaddr) {
+				dev_err(dev,
+					"Cannot memory-map BAR %d, aborting\n",
+					i);
+				return -ENODEV;
+			}
+		}
+
+		bars[j].bus_addr = pci_resource_start(pdev, i);
+		bars[j].res_index = i;
+		ionic->num_bars++;
+		j++;
+	}
+
+	return 0;
+}
+
+static void ionic_unmap_bars(struct ionic *ionic)
+{
+	struct ionic_dev_bar *bars = ionic->bars;
+	unsigned int i;
+
+	for (i = 0; i < IONIC_BARS_MAX; i++) {
+		if (bars[i].vaddr) {
+			iounmap(bars[i].vaddr);
+			bars[i].bus_addr = 0;
+			bars[i].vaddr = NULL;
+			bars[i].len = 0;
+		}
+	}
+}
+
 static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	struct device *dev = &pdev->dev;
 	struct ionic *ionic;
+	int err;
 
 	ionic = devm_kzalloc(dev, sizeof(*ionic), GFP_KERNEL);
 	if (!ionic)
 		return -ENOMEM;
 
 	ionic->pdev = pdev;
-	pci_set_drvdata(pdev, ionic);
 	ionic->dev = dev;
+	pci_set_drvdata(pdev, ionic);
+	mutex_init(&ionic->dev_cmd_lock);
+
+	/* Query system for DMA addressing limitation for the device. */
+	err = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(IONIC_ADDR_LEN));
+	if (err) {
+		dev_err(dev, "Unable to obtain 64-bit DMA for consistent allocations, aborting.  err=%d\n",
+			err);
+		goto err_out_clear_drvdata;
+	}
+
+	ionic_debugfs_add_dev(ionic);
+
+	/* Setup PCI device */
+	err = pci_enable_device_mem(pdev);
+	if (err) {
+		dev_err(dev, "Cannot enable PCI device: %d, aborting\n", err);
+		goto err_out_debugfs_del_dev;
+	}
+
+	err = pci_request_regions(pdev, DRV_NAME);
+	if (err) {
+		dev_err(dev, "Cannot request PCI regions: %d, aborting\n", err);
+		goto err_out_pci_disable_device;
+	}
+
+	pci_set_master(pdev);
+
+	err = ionic_map_bars(ionic);
+	if (err)
+		goto err_out_pci_clear_master;
+
+	/* Configure the device */
+	err = ionic_setup(ionic);
+	if (err) {
+		dev_err(dev, "Cannot setup device: %d, aborting\n", err);
+		goto err_out_unmap_bars;
+	}
+
+	err = ionic_identify(ionic);
+	if (err) {
+		dev_err(dev, "Cannot identify device: %d, aborting\n", err);
+		goto err_out_teardown;
+	}
+
+	err = ionic_init(ionic);
+	if (err) {
+		dev_err(dev, "Cannot init device: %d, aborting\n", err);
+		goto err_out_teardown;
+	}
 
 	return 0;
+
+err_out_teardown:
+	ionic_dev_teardown(ionic);
+err_out_unmap_bars:
+	ionic_unmap_bars(ionic);
+	pci_release_regions(pdev);
+err_out_pci_clear_master:
+	pci_clear_master(pdev);
+err_out_pci_disable_device:
+	pci_disable_device(pdev);
+err_out_debugfs_del_dev:
+	ionic_debugfs_del_dev(ionic);
+err_out_clear_drvdata:
+	mutex_destroy(&ionic->dev_cmd_lock);
+	pci_set_drvdata(pdev, NULL);
+
+	return err;
 }
 
 static void ionic_remove(struct pci_dev *pdev)
 {
 	struct ionic *ionic = pci_get_drvdata(pdev);
 
-	devm_kfree(&pdev->dev, ionic);
+	if (ionic) {
+		ionic_reset(ionic);
+		ionic_dev_teardown(ionic);
+		ionic_unmap_bars(ionic);
+		pci_release_regions(pdev);
+		pci_clear_master(pdev);
+		pci_disable_sriov(pdev);
+		pci_disable_device(pdev);
+		ionic_debugfs_del_dev(ionic);
+		mutex_destroy(&ionic->dev_cmd_lock);
+
+		devm_kfree(&pdev->dev, ionic);
+	}
 }
 
 static struct pci_driver ionic_driver = {
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_debugfs.c b/drivers/net/ethernet/pensando/ionic/ionic_debugfs.c
new file mode 100644
index 000000000000..c33a8bfcbcd4
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_debugfs.c
@@ -0,0 +1,67 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright(c) 2017 - 2019 Pensando Systems, Inc */
+
+#include <linux/netdevice.h>
+
+#include "ionic.h"
+#include "ionic_bus.h"
+#include "ionic_debugfs.h"
+
+#ifdef CONFIG_DEBUG_FS
+
+static struct dentry *ionic_dir;
+
+void ionic_debugfs_create(void)
+{
+	ionic_dir = debugfs_create_dir(DRV_NAME, NULL);
+}
+
+void ionic_debugfs_destroy(void)
+{
+	debugfs_remove_recursive(ionic_dir);
+}
+
+int ionic_debugfs_add_dev(struct ionic *ionic)
+{
+	struct dentry *dentry;
+
+	dentry = debugfs_create_dir(ionic_bus_info(ionic), ionic_dir);
+	if (IS_ERR_OR_NULL(dentry))
+		return PTR_ERR(dentry);
+
+	ionic->dentry = dentry;
+
+	return 0;
+}
+
+void ionic_debugfs_del_dev(struct ionic *ionic)
+{
+	debugfs_remove_recursive(ionic->dentry);
+	ionic->dentry = NULL;
+}
+
+static int identity_show(struct seq_file *seq, void *v)
+{
+	struct ionic *ionic = seq->private;
+	struct identity *ident = &ionic->ident;
+
+	seq_printf(seq, "nlifs:            %d\n", ident->dev.nlifs);
+	seq_printf(seq, "nintrs:           %d\n", ident->dev.nintrs);
+	seq_printf(seq, "ndbpgs_per_lif:   %d\n", ident->dev.ndbpgs_per_lif);
+	seq_printf(seq, "intr_coal_mult:   %d\n", ident->dev.intr_coal_mult);
+	seq_printf(seq, "intr_coal_div:    %d\n", ident->dev.intr_coal_div);
+
+	seq_printf(seq, "max_ucast_filters:  %d\n", ident->lif.eth.max_ucast_filters);
+	seq_printf(seq, "max_mcast_filters:  %d\n", ident->lif.eth.max_mcast_filters);
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(identity);
+
+int ionic_debugfs_add_ident(struct ionic *ionic)
+{
+	return debugfs_create_file("identity", 0400, ionic->dentry,
+				   ionic, &identity_fops) ? 0 : -ENOTSUPP;
+}
+
+#endif
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_debugfs.h b/drivers/net/ethernet/pensando/ionic/ionic_debugfs.h
new file mode 100644
index 000000000000..c3958b0c15b1
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_debugfs.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright(c) 2017 - 2019 Pensando Systems, Inc */
+
+#ifndef _IONIC_DEBUGFS_H_
+#define _IONIC_DEBUGFS_H_
+
+#include <linux/debugfs.h>
+
+#ifdef CONFIG_DEBUG_FS
+
+void ionic_debugfs_create(void);
+void ionic_debugfs_destroy(void);
+int ionic_debugfs_add_dev(struct ionic *ionic);
+void ionic_debugfs_del_dev(struct ionic *ionic);
+int ionic_debugfs_add_bars(struct ionic *ionic);
+int ionic_debugfs_add_dev_cmd(struct ionic *ionic);
+int ionic_debugfs_add_ident(struct ionic *ionic);
+#else
+static inline void ionic_debugfs_create(void) { }
+static inline void ionic_debugfs_destroy(void) { }
+static inline int ionic_debugfs_add_dev(struct ionic *ionic) { return 0; }
+static inline void ionic_debugfs_del_dev(struct ionic *ionic) { }
+static inline int ionic_debugfs_add_bars(struct ionic *ionic) { return 0; }
+static inline int ionic_debugfs_add_dev_cmd(struct ionic *ionic) { return 0; }
+static inline int ionic_debugfs_add_ident(struct ionic *ionic) { return 0; }
+#endif
+
+#endif /* _IONIC_DEBUGFS_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.c b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
new file mode 100644
index 000000000000..ca2ef0a1f620
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
@@ -0,0 +1,132 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright(c) 2017 - 2019 Pensando Systems, Inc */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/etherdevice.h>
+#include "ionic.h"
+#include "ionic_dev.h"
+
+void ionic_init_devinfo(struct ionic_dev *idev)
+{
+	idev->dev_info.asic_type = ioread8(&idev->dev_info_regs->asic_type);
+	idev->dev_info.asic_rev = ioread8(&idev->dev_info_regs->asic_rev);
+
+	memcpy_fromio(idev->dev_info.fw_version,
+		      idev->dev_info_regs->fw_version,
+		      IONIC_DEVINFO_FWVERS_BUFLEN);
+
+	memcpy_fromio(idev->dev_info.serial_num,
+		      idev->dev_info_regs->serial_num,
+		      IONIC_DEVINFO_SERIAL_BUFLEN);
+
+	idev->dev_info.fw_version[IONIC_DEVINFO_FWVERS_BUFLEN] = 0;
+	idev->dev_info.serial_num[IONIC_DEVINFO_SERIAL_BUFLEN] = 0;
+}
+
+int ionic_dev_setup(struct ionic *ionic)
+{
+	struct ionic_dev_bar *bar = ionic->bars;
+	unsigned int num_bars = ionic->num_bars;
+	struct ionic_dev *idev = &ionic->idev;
+	struct device *dev = ionic->dev;
+	u32 sig;
+
+	/* BAR0: dev_cmd and interrupts */
+	if (num_bars < 1) {
+		dev_err(dev, "No bars found, aborting\n");
+		return -EFAULT;
+	}
+
+	if (bar->len < BAR0_SIZE) {
+		dev_err(dev, "Resource bar size %lu too small, aborting\n",
+			bar->len);
+		return -EFAULT;
+	}
+
+	idev->dev_info_regs = bar->vaddr + BAR0_DEV_INFO_REGS_OFFSET;
+	idev->dev_cmd_regs = bar->vaddr + BAR0_DEV_CMD_REGS_OFFSET;
+	idev->intr_status = bar->vaddr + BAR0_INTR_STATUS_OFFSET;
+	idev->intr_ctrl = bar->vaddr + BAR0_INTR_CTRL_OFFSET;
+
+	sig = ioread32(&idev->dev_info_regs->signature);
+	if (sig != IONIC_DEV_INFO_SIGNATURE) {
+		dev_err(dev, "Incompatible firmware signature %x", sig);
+		return -EFAULT;
+	}
+
+	ionic_init_devinfo(idev);
+
+	/* BAR1: doorbells */
+	bar++;
+	if (num_bars < 2) {
+		dev_err(dev, "Doorbell bar missing, aborting\n");
+		return -EFAULT;
+	}
+
+	idev->db_pages = bar->vaddr;
+	idev->phy_db_pages = bar->bus_addr;
+
+	return 0;
+}
+
+void ionic_dev_teardown(struct ionic *ionic)
+{
+	/* place holder */
+}
+
+/* Devcmd Interface */
+u8 ionic_dev_cmd_status(struct ionic_dev *idev)
+{
+	return ioread8(&idev->dev_cmd_regs->comp.comp.status);
+}
+
+bool ionic_dev_cmd_done(struct ionic_dev *idev)
+{
+	return ioread32(&idev->dev_cmd_regs->done) & DEV_CMD_DONE;
+}
+
+void ionic_dev_cmd_comp(struct ionic_dev *idev, union dev_cmd_comp *comp)
+{
+	memcpy_fromio(comp, &idev->dev_cmd_regs->comp, sizeof(*comp));
+}
+
+void ionic_dev_cmd_go(struct ionic_dev *idev, union dev_cmd *cmd)
+{
+	memcpy_toio(&idev->dev_cmd_regs->cmd, cmd, sizeof(*cmd));
+	iowrite32(0, &idev->dev_cmd_regs->done);
+	iowrite32(1, &idev->dev_cmd_regs->doorbell);
+}
+
+/* Device commands */
+void ionic_dev_cmd_identify(struct ionic_dev *idev, u8 ver)
+{
+	union dev_cmd cmd = {
+		.identify.opcode = CMD_OPCODE_IDENTIFY,
+		.identify.ver = ver,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_init(struct ionic_dev *idev)
+{
+	union dev_cmd cmd = {
+		.init.opcode = CMD_OPCODE_INIT,
+		.init.type = 0,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_reset(struct ionic_dev *idev)
+{
+	union dev_cmd cmd = {
+		.reset.opcode = CMD_OPCODE_RESET,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.h b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
new file mode 100644
index 000000000000..66afab3ee396
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
@@ -0,0 +1,144 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright(c) 2017 - 2019 Pensando Systems, Inc */
+
+#ifndef _IONIC_DEV_H_
+#define _IONIC_DEV_H_
+
+#include <linux/mutex.h>
+#include <linux/workqueue.h>
+
+#include "ionic_if.h"
+#include "ionic_regs.h"
+
+struct ionic_dev_bar {
+	void __iomem *vaddr;
+	phys_addr_t bus_addr;
+	unsigned long len;
+	int res_index;
+};
+
+static inline void ionic_struct_size_checks(void)
+{
+	/* Registers */
+	BUILD_BUG_ON(sizeof(struct ionic_intr) != 32);
+
+	BUILD_BUG_ON(sizeof(struct doorbell) != 8);
+	BUILD_BUG_ON(sizeof(struct intr_ctrl) != 32);
+	BUILD_BUG_ON(sizeof(struct intr_status) != 8);
+
+	BUILD_BUG_ON(sizeof(union dev_regs) != 4096);
+	BUILD_BUG_ON(sizeof(union dev_info_regs) != 2048);
+	BUILD_BUG_ON(sizeof(union dev_cmd_regs) != 2048);
+
+	BUILD_BUG_ON(sizeof(struct lif_stats) != 1024);
+
+	BUILD_BUG_ON(sizeof(struct admin_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct admin_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct nop_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct nop_comp) != 16);
+
+	/* Device commands */
+	BUILD_BUG_ON(sizeof(struct dev_identify_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct dev_identify_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct dev_init_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct dev_init_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct dev_reset_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct dev_reset_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct dev_getattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct dev_getattr_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct dev_setattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct dev_setattr_comp) != 16);
+
+	/* Port commands */
+	BUILD_BUG_ON(sizeof(struct port_identify_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct port_identify_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct port_init_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct port_init_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct port_reset_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct port_reset_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct port_getattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct port_getattr_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct port_setattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct port_setattr_comp) != 16);
+
+	/* LIF commands */
+	BUILD_BUG_ON(sizeof(struct lif_init_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct lif_init_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct lif_reset_cmd) != 64);
+	BUILD_BUG_ON(sizeof(lif_reset_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct lif_getattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct lif_getattr_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct lif_setattr_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct lif_setattr_comp) != 16);
+
+	BUILD_BUG_ON(sizeof(struct q_init_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct q_init_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct q_control_cmd) != 64);
+	BUILD_BUG_ON(sizeof(q_control_comp) != 16);
+
+	BUILD_BUG_ON(sizeof(struct rx_mode_set_cmd) != 64);
+	BUILD_BUG_ON(sizeof(rx_mode_set_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct rx_filter_add_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct rx_filter_add_comp) != 16);
+	BUILD_BUG_ON(sizeof(struct rx_filter_del_cmd) != 64);
+	BUILD_BUG_ON(sizeof(rx_filter_del_comp) != 16);
+
+	/* RDMA commands */
+	BUILD_BUG_ON(sizeof(struct rdma_reset_cmd) != 64);
+	BUILD_BUG_ON(sizeof(struct rdma_queue_cmd) != 64);
+
+	/* Events */
+	BUILD_BUG_ON(sizeof(struct notifyq_cmd) != 4);
+	BUILD_BUG_ON(sizeof(union notifyq_comp) != 64);
+	BUILD_BUG_ON(sizeof(struct notifyq_event) != 64);
+	BUILD_BUG_ON(sizeof(struct link_change_event) != 64);
+	BUILD_BUG_ON(sizeof(struct reset_event) != 64);
+	BUILD_BUG_ON(sizeof(struct heartbeat_event) != 64);
+	BUILD_BUG_ON(sizeof(struct log_event) != 64);
+
+	/* I/O */
+	BUILD_BUG_ON(sizeof(struct txq_desc) != 16);
+	BUILD_BUG_ON(sizeof(struct txq_sg_desc) != 128);
+	BUILD_BUG_ON(sizeof(struct txq_comp) != 16);
+
+	BUILD_BUG_ON(sizeof(struct rxq_desc) != 16);
+	BUILD_BUG_ON(sizeof(struct rxq_sg_desc) != 128);
+	BUILD_BUG_ON(sizeof(struct rxq_comp) != 16);
+}
+
+struct ionic_devinfo {
+	u8 asic_type;
+	u8 asic_rev;
+	char fw_version[IONIC_DEVINFO_FWVERS_BUFLEN + 1];
+	char serial_num[IONIC_DEVINFO_SERIAL_BUFLEN + 1];
+};
+
+struct ionic_dev {
+	union dev_info_regs __iomem *dev_info_regs;
+	union dev_cmd_regs __iomem *dev_cmd_regs;
+
+	u64 __iomem *db_pages;
+	dma_addr_t phy_db_pages;
+
+	struct ionic_intr __iomem *intr_ctrl;
+	u64 __iomem *intr_status;
+
+	struct ionic_devinfo dev_info;
+};
+
+struct ionic;
+
+void ionic_init_devinfo(struct ionic_dev *idev);
+int ionic_dev_setup(struct ionic *ionic);
+void ionic_dev_teardown(struct ionic *ionic);
+
+void ionic_dev_cmd_go(struct ionic_dev *idev, union dev_cmd *cmd);
+u8 ionic_dev_cmd_status(struct ionic_dev *idev);
+bool ionic_dev_cmd_done(struct ionic_dev *idev);
+void ionic_dev_cmd_comp(struct ionic_dev *idev, union dev_cmd_comp *comp);
+
+void ionic_dev_cmd_identify(struct ionic_dev *idev, u8 ver);
+void ionic_dev_cmd_init(struct ionic_dev *idev);
+void ionic_dev_cmd_reset(struct ionic_dev *idev);
+
+#endif /* _IONIC_DEV_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_if.h b/drivers/net/ethernet/pensando/ionic/ionic_if.h
new file mode 100644
index 000000000000..0afbe842c81e
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_if.h
@@ -0,0 +1,2552 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB OR BSD-2-Clause */
+/* Copyright (c) 2017-2019 Pensando Systems, Inc.  All rights reserved. */
+
+#ifndef _IONIC_IF_H_
+#define _IONIC_IF_H_
+
+#pragma pack(push, 1)
+
+#define IONIC_DEV_INFO_SIGNATURE		0x44455649      /* 'DEVI' */
+#define IONIC_DEV_INFO_VERSION			1
+#define IONIC_IFNAMSIZ				16
+
+/**
+ * Commands
+ */
+enum cmd_opcode {
+	CMD_OPCODE_NOP				= 0,
+
+	/* Device commands */
+	CMD_OPCODE_IDENTIFY			= 1,
+	CMD_OPCODE_INIT				= 2,
+	CMD_OPCODE_RESET			= 3,
+	CMD_OPCODE_GETATTR			= 4,
+	CMD_OPCODE_SETATTR			= 5,
+
+	/* Port commands */
+	CMD_OPCODE_PORT_IDENTIFY		= 10,
+	CMD_OPCODE_PORT_INIT			= 11,
+	CMD_OPCODE_PORT_RESET			= 12,
+	CMD_OPCODE_PORT_GETATTR			= 13,
+	CMD_OPCODE_PORT_SETATTR			= 14,
+
+	/* LIF commands */
+	CMD_OPCODE_LIF_IDENTIFY			= 20,
+	CMD_OPCODE_LIF_INIT			= 21,
+	CMD_OPCODE_LIF_RESET			= 22,
+	CMD_OPCODE_LIF_GETATTR			= 23,
+	CMD_OPCODE_LIF_SETATTR			= 24,
+
+	CMD_OPCODE_RX_MODE_SET			= 30,
+	CMD_OPCODE_RX_FILTER_ADD		= 31,
+	CMD_OPCODE_RX_FILTER_DEL		= 32,
+
+	/* Queue commands */
+	CMD_OPCODE_Q_INIT			= 40,
+	CMD_OPCODE_Q_CONTROL			= 41,
+
+	/* RDMA commands */
+	CMD_OPCODE_RDMA_RESET_LIF		= 50,
+	CMD_OPCODE_RDMA_CREATE_EQ		= 51,
+	CMD_OPCODE_RDMA_CREATE_CQ		= 52,
+	CMD_OPCODE_RDMA_CREATE_ADMINQ		= 53,
+
+	/* QoS commands */
+	CMD_OPCODE_QOS_CLASS_IDENTIFY		= 240,
+	CMD_OPCODE_QOS_CLASS_INIT		= 241,
+	CMD_OPCODE_QOS_CLASS_RESET		= 242,
+
+	/* Firmware commands */
+	CMD_OPCODE_FW_DOWNLOAD			= 254,
+	CMD_OPCODE_FW_CONTROL			= 255,
+};
+
+/**
+ * Command Return codes
+ */
+enum status_code {
+	IONIC_RC_SUCCESS	= 0,	/* Success */
+	IONIC_RC_EVERSION	= 1,	/* Incorrect version for request */
+	IONIC_RC_EOPCODE	= 2,	/* Invalid cmd opcode */
+	IONIC_RC_EIO		= 3,	/* I/O error */
+	IONIC_RC_EPERM		= 4,	/* Permission denied */
+	IONIC_RC_EQID		= 5,	/* Bad qid */
+	IONIC_RC_EQTYPE		= 6,	/* Bad qtype */
+	IONIC_RC_ENOENT		= 7,	/* No such element */
+	IONIC_RC_EINTR		= 8,	/* operation interrupted */
+	IONIC_RC_EAGAIN		= 9,	/* Try again */
+	IONIC_RC_ENOMEM		= 10,	/* Out of memory */
+	IONIC_RC_EFAULT		= 11,	/* Bad address */
+	IONIC_RC_EBUSY		= 12,	/* Device or resource busy */
+	IONIC_RC_EEXIST		= 13,	/* object already exists */
+	IONIC_RC_EINVAL		= 14,	/* Invalid argument */
+	IONIC_RC_ENOSPC		= 15,	/* No space left or alloc failure */
+	IONIC_RC_ERANGE		= 16,	/* Parameter out of range */
+	IONIC_RC_BAD_ADDR	= 17,	/* Descriptor contains a bad ptr */
+	IONIC_RC_DEV_CMD	= 18,	/* Device cmd attempted on AdminQ */
+	IONIC_RC_ERROR		= 29,	/* Generic error */
+
+	IONIC_RC_ERDMA		= 30,	/* Generic RDMA error */
+};
+
+enum notifyq_opcode {
+	EVENT_OPCODE_LINK_CHANGE	= 1,
+	EVENT_OPCODE_RESET		= 2,
+	EVENT_OPCODE_HEARTBEAT		= 3,
+	EVENT_OPCODE_LOG		= 4,
+};
+
+/**
+ * struct cmd - General admin command format
+ * @opcode:     Opcode for the command
+ * @lif_index:  LIF index
+ * @cmd_data:   Opcode-specific command bytes
+ */
+struct admin_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	u8     cmd_data[60];
+};
+
+/**
+ * struct admin_comp - General admin command completion format
+ * @status:     The status of the command (enum status_code)
+ * @comp_index: The index in the descriptor ring for which this
+ *              is the completion.
+ * @cmd_data:   Command-specific bytes.
+ * @color:      Color bit.  (Always 0 for commands issued to the
+ *              Device Cmd Registers.)
+ */
+struct admin_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	u8     cmd_data[11];
+	u8     color;
+#define IONIC_COMP_COLOR_MASK  0x80
+};
+
+static inline u8 color_match(u8 color, u8 done_color)
+{
+	return (!!(color & IONIC_COMP_COLOR_MASK)) == done_color;
+}
+
+/**
+ * struct nop_cmd - NOP command
+ * @opcode: opcode
+ */
+struct nop_cmd {
+	u8 opcode;
+	u8 rsvd[63];
+};
+
+/**
+ * struct nop_comp - NOP command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct nop_comp {
+	u8 status;
+	u8 rsvd[15];
+};
+
+/**
+ * struct dev_init_cmd - Device init command
+ * @opcode:    opcode
+ * @type:      device type
+ */
+struct dev_init_cmd {
+	u8     opcode;
+	u8     type;
+	u8     rsvd[62];
+};
+
+/**
+ * struct init_comp - Device init command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct dev_init_comp {
+	u8 status;
+	u8 rsvd[15];
+};
+
+/**
+ * struct dev_reset_cmd - Device reset command
+ * @opcode: opcode
+ */
+struct dev_reset_cmd {
+	u8 opcode;
+	u8 rsvd[63];
+};
+
+/**
+ * struct reset_comp - Reset command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct dev_reset_comp {
+	u8 status;
+	u8 rsvd[15];
+};
+
+#define IONIC_IDENTITY_VERSION_1	1
+
+/**
+ * struct dev_identify_cmd - Driver/device identify command
+ * @opcode:  opcode
+ * @ver:     Highest version of identify supported by driver
+ */
+struct dev_identify_cmd {
+	u8 opcode;
+	u8 ver;
+	u8 rsvd[62];
+};
+
+/**
+ * struct dev_identify_comp - Driver/device identify command completion
+ * @status: The status of the command (enum status_code)
+ * @ver:    Version of identify returned by device
+ */
+struct dev_identify_comp {
+	u8 status;
+	u8 ver;
+	u8 rsvd[14];
+};
+
+enum os_type {
+	IONIC_OS_TYPE_LINUX   = 1,
+	IONIC_OS_TYPE_WIN     = 2,
+	IONIC_OS_TYPE_DPDK    = 3,
+	IONIC_OS_TYPE_FREEBSD = 4,
+	IONIC_OS_TYPE_IPXE    = 5,
+	IONIC_OS_TYPE_ESXI    = 6,
+};
+
+/**
+ * union drv_identity - driver identity information
+ * @os_type:          OS type (see enum os_type)
+ * @os_dist:          OS distribution, numeric format
+ * @os_dist_str:      OS distribution, string format
+ * @kernel_ver:       Kernel version, numeric format
+ * @kernel_ver_str:   Kernel version, string format
+ * @driver_ver_str:   Driver version, string format
+ */
+union drv_identity {
+	struct {
+		__le32 os_type;
+		__le32 os_dist;
+		char   os_dist_str[128];
+		__le32 kernel_ver;
+		char   kernel_ver_str[32];
+		char   driver_ver_str[32];
+	};
+	__le32 words[512];
+};
+
+/**
+ * union dev_identity - device identity information
+ * @version:          Version of device identify
+ * @type:             Identify type (0 for now)
+ * @nports:           Number of ports provisioned
+ * @nlifs:            Number of LIFs provisioned
+ * @nintrs:           Number of interrupts provisioned
+ * @ndbpgs_per_lif:   Number of doorbell pages per LIF
+ * @intr_coal_mult:   Interrupt coalescing multiplication factor.
+ *                    Scale user-supplied interrupt coalescing
+ *                    value in usecs to device units using:
+ *                    device units = usecs * mult / div
+ * @intr_coal_div:    Interrupt coalescing division factor.
+ *                    Scale user-supplied interrupt coalescing
+ *                    value in usecs to device units using:
+ *                    device units = usecs * mult / div
+ *
+ */
+union dev_identity {
+	struct {
+		u8     version;
+		u8     type;
+		u8     rsvd[2];
+		u8     nports;
+		u8     rsvd2[3];
+		__le32 nlifs;
+		__le32 nintrs;
+		__le32 ndbpgs_per_lif;
+		__le32 intr_coal_mult;
+		__le32 intr_coal_div;
+	};
+	__le32 words[512];
+};
+
+enum lif_type {
+	IONIC_LIF_TYPE_CLASSIC = 0,
+	IONIC_LIF_TYPE_MACVLAN = 1,
+	IONIC_LIF_TYPE_NETQUEUE = 2,
+};
+
+/**
+ * struct lif_identify_cmd - lif identify command
+ * @opcode:  opcode
+ * @type:    lif type (enum lif_type)
+ * @ver:     version of identify returned by device
+ */
+struct lif_identify_cmd {
+	u8 opcode;
+	u8 type;
+	u8 ver;
+	u8 rsvd[61];
+};
+
+/**
+ * struct lif_identify_comp - lif identify command completion
+ * @status:  status of the command (enum status_code)
+ * @ver:     version of identify returned by device
+ */
+struct lif_identify_comp {
+	u8 status;
+	u8 ver;
+	u8 rsvd2[14];
+};
+
+enum lif_capability {
+	IONIC_LIF_CAP_ETH        = BIT(0),
+	IONIC_LIF_CAP_RDMA       = BIT(1),
+};
+
+/**
+ * Logical Queue Types
+ */
+enum logical_qtype {
+	IONIC_QTYPE_ADMINQ  = 0,
+	IONIC_QTYPE_NOTIFYQ = 1,
+	IONIC_QTYPE_RXQ     = 2,
+	IONIC_QTYPE_TXQ     = 3,
+	IONIC_QTYPE_EQ      = 4,
+	IONIC_QTYPE_MAX     = 16,
+};
+
+/**
+ * struct lif_logical_qtype - Descriptor of logical to hardware queue type.
+ * @qtype:          Hardware Queue Type.
+ * @qid_count:      Number of Queue IDs of the logical type.
+ * @qid_base:       Minimum Queue ID of the logical type.
+ */
+struct lif_logical_qtype {
+	u8     qtype;
+	u8     rsvd[3];
+	__le32 qid_count;
+	__le32 qid_base;
+};
+
+enum lif_state {
+	IONIC_LIF_DISABLE	= 0,
+	IONIC_LIF_ENABLE	= 1,
+	IONIC_LIF_HANG_RESET	= 2,
+};
+
+/**
+ * LIF configuration
+ * @state:          lif state (enum lif_state)
+ * @name:           lif name
+ * @mtu:            mtu
+ * @mac:            station mac address
+ * @features:       features (enum eth_hw_features)
+ * @queue_count:    queue counts per queue-type
+ */
+union lif_config {
+	struct {
+		u8     state;
+		u8     rsvd[3];
+		char   name[IONIC_IFNAMSIZ];
+		__le32 mtu;
+		u8     mac[6];
+		u8     rsvd2[2];
+		__le64 features;
+		__le32 queue_count[IONIC_QTYPE_MAX];
+	};
+	__le32 words[64];
+};
+
+/**
+ * struct lif_identity - lif identity information (type-specific)
+ *
+ * @capabilities    LIF capabilities
+ *
+ * Ethernet:
+ *     @version:          Ethernet identify structure version.
+ *     @features:         Ethernet features supported on this lif type.
+ *     @max_ucast_filters:  Number of perfect unicast addresses supported.
+ *     @max_mcast_filters:  Number of perfect multicast addresses supported.
+ *     @min_frame_size:   Minimum size of frames to be sent
+ *     @max_frame_size:   Maximim size of frames to be sent
+ *     @config:           LIF config struct with features, mtu, mac, q counts
+ *
+ * RDMA:
+ *     @version:         RDMA version of opcodes and queue descriptors.
+ *     @qp_opcodes:      Number of rdma queue pair opcodes supported.
+ *     @admin_opcodes:   Number of rdma admin opcodes supported.
+ *     @npts_per_lif:    Page table size per lif
+ *     @nmrs_per_lif:    Number of memory regions per lif
+ *     @nahs_per_lif:    Number of address handles per lif
+ *     @max_stride:      Max work request stride.
+ *     @cl_stride:       Cache line stride.
+ *     @pte_stride:      Page table entry stride.
+ *     @rrq_stride:      Remote RQ work request stride.
+ *     @rsq_stride:      Remote SQ work request stride.
+ *     @dcqcn_profiles:  Number of DCQCN profiles
+ *     @aq_qtype:        RDMA Admin Qtype.
+ *     @sq_qtype:        RDMA Send Qtype.
+ *     @rq_qtype:        RDMA Receive Qtype.
+ *     @cq_qtype:        RDMA Completion Qtype.
+ *     @eq_qtype:        RDMA Event Qtype.
+ */
+union lif_identity {
+	struct {
+		__le64 capabilities;
+
+		struct {
+			u8 version;
+			u8 rsvd[3];
+			__le32 max_ucast_filters;
+			__le32 max_mcast_filters;
+			__le16 rss_ind_tbl_sz;
+			__le32 min_frame_size;
+			__le32 max_frame_size;
+			u8 rsvd2[106];
+			union lif_config config;
+		} eth;
+
+		struct {
+			u8 version;
+			u8 qp_opcodes;
+			u8 admin_opcodes;
+			u8 rsvd;
+			__le32 npts_per_lif;
+			__le32 nmrs_per_lif;
+			__le32 nahs_per_lif;
+			u8 max_stride;
+			u8 cl_stride;
+			u8 pte_stride;
+			u8 rrq_stride;
+			u8 rsq_stride;
+			u8 dcqcn_profiles;
+			u8 rsvd_dimensions[10];
+			struct lif_logical_qtype aq_qtype;
+			struct lif_logical_qtype sq_qtype;
+			struct lif_logical_qtype rq_qtype;
+			struct lif_logical_qtype cq_qtype;
+			struct lif_logical_qtype eq_qtype;
+		} rdma;
+	};
+	__le32 words[512];
+};
+
+/**
+ * struct lif_init_cmd - LIF init command
+ * @opcode:       opcode
+ * @type:         LIF type (enum lif_type)
+ * @index:        LIF index
+ * @info_pa:      destination address for lif info (struct lif_info)
+ */
+struct lif_init_cmd {
+	u8     opcode;
+	u8     type;
+	__le16 index;
+	__le32 rsvd;
+	__le64 info_pa;
+	u8     rsvd2[48];
+};
+
+/**
+ * struct lif_init_comp - LIF init command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct lif_init_comp {
+	u8 status;
+	u8 rsvd;
+	__le16 hw_index;
+	u8 rsvd2[12];
+};
+
+/**
+ * struct q_init_cmd - Queue init command
+ * @opcode:       opcode
+ * @type:         Logical queue type
+ * @ver:          Queue version (defines opcode/descriptor scope)
+ * @lif_index:    LIF index
+ * @index:        (lif, qtype) relative admin queue index
+ * @intr_index:   Interrupt control register index
+ * @pid:          Process ID
+ * @flags:
+ *    IRQ:        Interrupt requested on completion
+ *    ENA:        Enable the queue.  If ENA=0 the queue is initialized
+ *                but remains disabled, to be later enabled with the
+ *                Queue Enable command.  If ENA=1, then queue is
+ *                initialized and then enabled.
+ *    SG:         Enable Scatter-Gather on the queue.
+ *                in number of descs.  The actual ring size is
+ *                (1 << ring_size).  For example, to
+ *                select a ring size of 64 descriptors write
+ *                ring_size = 6.  The minimum ring_size value is 2
+ *                for a ring size of 4 descriptors.  The maximum
+ *                ring_size value is 16 for a ring size of 64k
+ *                descriptors.  Values of ring_size <2 and >16 are
+ *                reserved.
+ *    EQ:         Enable the Event Queue
+ * @cos:          Class of service for this queue.
+ * @ring_size:    Queue ring size, encoded as a log2(size)
+ * @ring_base:    Queue ring base address
+ * @cq_ring_base: Completion queue ring base address
+ * @sg_ring_base: Scatter/Gather ring base address
+ * @eq_index:	  Event queue index
+ */
+struct q_init_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	u8     type;
+	u8     ver;
+	u8     rsvd1[2];
+	__le32 index;
+	__le16 pid;
+	__le16 intr_index;
+	__le16 flags;
+#define IONIC_QINIT_F_IRQ	0x01	/* Request interrupt on completion */
+#define IONIC_QINIT_F_ENA	0x02	/* Enable the queue */
+#define IONIC_QINIT_F_SG	0x04	/* Enable scatter/gather on the queue */
+#define IONIC_QINIT_F_EQ	0x08	/* Enable event queue */
+#define IONIC_QINIT_F_DEBUG 0x80	/* Enable queue debugging */
+	u8     cos;
+	u8     ring_size;
+	__le64 ring_base;
+	__le64 cq_ring_base;
+	__le64 sg_ring_base;
+	__le32 eq_index;
+	u8     rsvd2[16];
+};
+
+/**
+ * struct q_init_comp - Queue init command completion
+ * @status:     The status of the command (enum status_code)
+ * @ver:        Queue version (defines opcode/descriptor scope)
+ * @comp_index: The index in the descriptor ring for which this
+ *              is the completion.
+ * @hw_index:   Hardware Queue ID
+ * @hw_type:    Hardware Queue type
+ * @color:      Color
+ */
+struct q_init_comp {
+	u8     status;
+	u8     ver;
+	__le16 comp_index;
+	__le32 hw_index;
+	u8     hw_type;
+	u8     rsvd2[6];
+	u8     color;
+};
+
+/* the device's internal addressing uses up to 52 bits */
+#define IONIC_ADDR_LEN		52
+#define IONIC_ADDR_MASK		(BIT_ULL(IONIC_ADDR_LEN) - 1)
+
+enum txq_desc_opcode {
+	IONIC_TXQ_DESC_OPCODE_CSUM_NONE = 0,
+	IONIC_TXQ_DESC_OPCODE_CSUM_PARTIAL = 1,
+	IONIC_TXQ_DESC_OPCODE_CSUM_HW = 2,
+	IONIC_TXQ_DESC_OPCODE_TSO = 3,
+};
+
+/**
+ * struct txq_desc - Ethernet Tx queue descriptor format
+ * @opcode:       Tx operation, see TXQ_DESC_OPCODE_*:
+ *
+ *                   IONIC_TXQ_DESC_OPCODE_CSUM_NONE:
+ *
+ *                      Non-offload send.  No segmentation,
+ *                      fragmentation or checksum calc/insertion is
+ *                      performed by device; packet is prepared
+ *                      to send by software stack and requires
+ *                      no further manipulation from device.
+ *
+ *                   IONIC_TXQ_DESC_OPCODE_CSUM_PARTIAL:
+ *
+ *                      Offload 16-bit L4 checksum
+ *                      calculation/insertion.  The device will
+ *                      calculate the L4 checksum value and
+ *                      insert the result in the packet's L4
+ *                      header checksum field.  The L4 checksum
+ *                      is calculated starting at @csum_start bytes
+ *                      into the packet to the end of the packet.
+ *                      The checksum insertion position is given
+ *                      in @csum_offset.  This feature is only
+ *                      applicable to protocols such as TCP, UDP
+ *                      and ICMP where a standard (i.e. the
+ *                      'IP-style' checksum) one's complement
+ *                      16-bit checksum is used, using an IP
+ *                      pseudo-header to seed the calculation.
+ *                      Software will preload the L4 checksum
+ *                      field with the IP pseudo-header checksum.
+ *
+ *                      For tunnel encapsulation, @csum_start and
+ *                      @csum_offset refer to the inner L4
+ *                      header.  Supported tunnels encapsulations
+ *                      are: IPIP, GRE, and UDP.  If the @encap
+ *                      is clear, no further processing by the
+ *                      device is required; software will
+ *                      calculate the outer header checksums.  If
+ *                      the @encap is set, the device will
+ *                      offload the outer header checksums using
+ *                      LCO (local checksum offload) (see
+ *                      Documentation/networking/checksum-
+ *                      offloads.txt for more info).
+ *
+ *                   IONIC_TXQ_DESC_OPCODE_CSUM_HW:
+ *
+ *                      Offload 16-bit checksum computation to hardware.
+ *                      If @csum_l3 is set then the packet's L3 checksum is
+ *                      updated. Similarly, if @csum_l4 is set the the L4
+ *                      checksum is updated. If @encap is set then encap header
+ *                      checksums are also updated.
+ *
+ *                   IONIC_TXQ_DESC_OPCODE_TSO:
+ *
+ *                      Device preforms TCP segmentation offload
+ *                      (TSO).  @hdr_len is the number of bytes
+ *                      to the end of TCP header (the offset to
+ *                      the TCP payload).  @mss is the desired
+ *                      MSS, the TCP payload length for each
+ *                      segment.  The device will calculate/
+ *                      insert IP (IPv4 only) and TCP checksums
+ *                      for each segment.  In the first data
+ *                      buffer containing the header template,
+ *                      the driver will set IPv4 checksum to 0
+ *                      and preload TCP checksum with the IP
+ *                      pseudo header calculated with IP length = 0.
+ *
+ *                      Supported tunnel encapsulations are IPIP,
+ *                      layer-3 GRE, and UDP. @hdr_len includes
+ *                      both outer and inner headers.  The driver
+ *                      will set IPv4 checksum to zero and
+ *                      preload TCP checksum with IP pseudo
+ *                      header on the inner header.
+ *
+ *                      TCP ECN offload is supported.  The device
+ *                      will set CWR flag in the first segment if
+ *                      CWR is set in the template header, and
+ *                      clear CWR in remaining segments.
+ * @flags:
+ *                vlan:
+ *                    Insert an L2 VLAN header using @vlan_tci.
+ *                encap:
+ *                    Calculate encap header checksum.
+ *                csum_l3:
+ *                    Compute L3 header checksum.
+ *                csum_l4:
+ *                    Compute L4 header checksum.
+ *                tso_sot:
+ *                    TSO start
+ *                tso_eot:
+ *                    TSO end
+ * @num_sg_elems: Number of scatter-gather elements in SG
+ *                descriptor
+ * @addr:         First data buffer's DMA address.
+ *                (Subsequent data buffers are on txq_sg_desc).
+ * @len:          First data buffer's length, in bytes
+ * @vlan_tci:     VLAN tag to insert in the packet (if requested
+ *                by @V-bit).  Includes .1p and .1q tags
+ * @hdr_len:      Length of packet headers, including
+ *                encapsulating outer header, if applicable.
+ *                Valid for opcodes TXQ_DESC_OPCODE_CALC_CSUM and
+ *                TXQ_DESC_OPCODE_TSO.  Should be set to zero for
+ *                all other modes.  For
+ *                TXQ_DESC_OPCODE_CALC_CSUM, @hdr_len is length
+ *                of headers up to inner-most L4 header.  For
+ *                TXQ_DESC_OPCODE_TSO, @hdr_len is up to
+ *                inner-most L4 payload, so inclusive of
+ *                inner-most L4 header.
+ * @mss:          Desired MSS value for TSO.  Only applicable for
+ *                TXQ_DESC_OPCODE_TSO.
+ * @csum_start:   Offset into inner-most L3 header of checksum
+ * @csum_offset:  Offset into inner-most L4 header of checksum
+ */
+
+#define IONIC_TXQ_DESC_OPCODE_MASK		0xf
+#define IONIC_TXQ_DESC_OPCODE_SHIFT		4
+#define IONIC_TXQ_DESC_FLAGS_MASK		0xf
+#define IONIC_TXQ_DESC_FLAGS_SHIFT		0
+#define IONIC_TXQ_DESC_NSGE_MASK		0xf
+#define IONIC_TXQ_DESC_NSGE_SHIFT		8
+#define IONIC_TXQ_DESC_ADDR_MASK		(BIT_ULL(IONIC_ADDR_LEN) - 1)
+#define IONIC_TXQ_DESC_ADDR_SHIFT		12
+
+/* common flags */
+#define IONIC_TXQ_DESC_FLAG_VLAN		0x1
+#define IONIC_TXQ_DESC_FLAG_ENCAP		0x2
+
+/* flags for csum_hw opcode */
+#define IONIC_TXQ_DESC_FLAG_CSUM_L3		0x4
+#define IONIC_TXQ_DESC_FLAG_CSUM_L4		0x8
+
+/* flags for tso opcode */
+#define IONIC_TXQ_DESC_FLAG_TSO_SOT		0x4
+#define IONIC_TXQ_DESC_FLAG_TSO_EOT		0x8
+
+struct txq_desc {
+	__le64  cmd;
+	__le16  len;
+	union {
+		__le16  vlan_tci;
+		__le16  hword0;
+	};
+	union {
+		__le16  csum_start;
+		__le16  hdr_len;
+		__le16  hword1;
+	};
+	union {
+		__le16  csum_offset;
+		__le16  mss;
+		__le16  hword2;
+	};
+};
+
+static inline u64 encode_txq_desc_cmd(u8 opcode, u8 flags,
+				      u8 nsge, u64 addr)
+{
+	u64 cmd;
+
+	cmd = (opcode & IONIC_TXQ_DESC_OPCODE_MASK) << IONIC_TXQ_DESC_OPCODE_SHIFT;
+	cmd |= (flags & IONIC_TXQ_DESC_FLAGS_MASK) << IONIC_TXQ_DESC_FLAGS_SHIFT;
+	cmd |= (nsge & IONIC_TXQ_DESC_NSGE_MASK) << IONIC_TXQ_DESC_NSGE_SHIFT;
+	cmd |= (addr & IONIC_TXQ_DESC_ADDR_MASK) << IONIC_TXQ_DESC_ADDR_SHIFT;
+
+	return cmd;
+};
+
+static inline void decode_txq_desc_cmd(u64 cmd, u8 *opcode, u8 *flags,
+				       u8 *nsge, u64 *addr)
+{
+	*opcode = (cmd >> IONIC_TXQ_DESC_OPCODE_SHIFT) & IONIC_TXQ_DESC_OPCODE_MASK;
+	*flags = (cmd >> IONIC_TXQ_DESC_FLAGS_SHIFT) & IONIC_TXQ_DESC_FLAGS_MASK;
+	*nsge = (cmd >> IONIC_TXQ_DESC_NSGE_SHIFT) & IONIC_TXQ_DESC_NSGE_MASK;
+	*addr = (cmd >> IONIC_TXQ_DESC_ADDR_SHIFT) & IONIC_TXQ_DESC_ADDR_MASK;
+};
+
+#define IONIC_TX_MAX_SG_ELEMS	8
+#define IONIC_RX_MAX_SG_ELEMS	8
+
+/**
+ * struct txq_sg_desc - Transmit scatter-gather (SG) list
+ * @addr:      DMA address of SG element data buffer
+ * @len:       Length of SG element data buffer, in bytes
+ */
+struct txq_sg_desc {
+	struct txq_sg_elem {
+		__le64 addr;
+		__le16 len;
+		__le16 rsvd[3];
+	} elems[IONIC_TX_MAX_SG_ELEMS];
+};
+
+/**
+ * struct txq_comp - Ethernet transmit queue completion descriptor
+ * @status:     The status of the command (enum status_code)
+ * @comp_index: The index in the descriptor ring for which this
+ *                 is the completion.
+ * @color:      Color bit.
+ */
+struct txq_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	u8     rsvd2[11];
+	u8     color;
+};
+
+enum rxq_desc_opcode {
+	RXQ_DESC_OPCODE_SIMPLE = 0,
+	RXQ_DESC_OPCODE_SG = 1,
+};
+
+/**
+ * struct rxq_desc - Ethernet Rx queue descriptor format
+ * @opcode:       Rx operation, see RXQ_DESC_OPCODE_*:
+ *
+ *                   RXQ_DESC_OPCODE_SIMPLE:
+ *
+ *                      Receive full packet into data buffer
+ *                      starting at @addr.  Results of
+ *                      receive, including actual bytes received,
+ *                      are recorded in Rx completion descriptor.
+ *
+ * @len:          Data buffer's length, in bytes.
+ * @addr:         Data buffer's DMA address
+ */
+struct rxq_desc {
+	u8     opcode;
+	u8     rsvd[5];
+	__le16 len;
+	__le64 addr;
+};
+
+/**
+ * struct rxq_sg_desc - Receive scatter-gather (SG) list
+ * @addr:      DMA address of SG element data buffer
+ * @len:       Length of SG element data buffer, in bytes
+ */
+struct rxq_sg_desc {
+	struct rxq_sg_elem {
+		__le64 addr;
+		__le16 len;
+		__le16 rsvd[3];
+	} elems[IONIC_RX_MAX_SG_ELEMS];
+};
+
+/**
+ * struct rxq_comp - Ethernet receive queue completion descriptor
+ * @status:       The status of the command (enum status_code)
+ * @num_sg_elems: Number of SG elements used by this descriptor
+ * @comp_index:   The index in the descriptor ring for which this
+ *                is the completion.
+ * @rss_hash:     32-bit RSS hash
+ * @csum:         16-bit sum of the packet's L2 payload.
+ *                If the packet's L2 payload is odd length, an extra
+ *                zero-value byte is included in the @csum calculation but
+ *                not included in @len.
+ * @vlan_tci:     VLAN tag stripped from the packet.  Valid if @VLAN is
+ *                set.  Includes .1p and .1q tags.
+ * @len:          Received packet length, in bytes.  Excludes FCS.
+ * @csum_calc     L2 payload checksum is computed or not
+ * @csum_tcp_ok:  The TCP checksum calculated by the device
+ *                matched the checksum in the receive packet's
+ *                TCP header
+ * @csum_tcp_bad: The TCP checksum calculated by the device did
+ *                not match the checksum in the receive packet's
+ *                TCP header.
+ * @csum_udp_ok:  The UDP checksum calculated by the device
+ *                matched the checksum in the receive packet's
+ *                UDP header
+ * @csum_udp_bad: The UDP checksum calculated by the device did
+ *                not match the checksum in the receive packet's
+ *                UDP header.
+ * @csum_ip_ok:   The IPv4 checksum calculated by the device
+ *                matched the checksum in the receive packet's
+ *                first IPv4 header.  If the receive packet
+ *                contains both a tunnel IPv4 header and a
+ *                transport IPv4 header, the device validates the
+ *                checksum for the both IPv4 headers.
+ * @csum_ip_bad:  The IPv4 checksum calculated by the device did
+ *                not match the checksum in the receive packet's
+ *                first IPv4 header. If the receive packet
+ *                contains both a tunnel IPv4 header and a
+ *                transport IPv4 header, the device validates the
+ *                checksum for both IP headers.
+ * @VLAN:         VLAN header was stripped and placed in @vlan_tci.
+ * @pkt_type:     Packet type
+ * @color:        Color bit.
+ */
+struct rxq_comp {
+	u8     status;
+	u8     num_sg_elems;
+	__le16 comp_index;
+	__le32 rss_hash;
+	__le16 csum;
+	__le16 vlan_tci;
+	__le16 len;
+	u8     csum_flags;
+#define IONIC_RXQ_COMP_CSUM_F_TCP_OK	0x01
+#define IONIC_RXQ_COMP_CSUM_F_TCP_BAD	0x02
+#define IONIC_RXQ_COMP_CSUM_F_UDP_OK	0x04
+#define IONIC_RXQ_COMP_CSUM_F_UDP_BAD	0x08
+#define IONIC_RXQ_COMP_CSUM_F_IP_OK	0x10
+#define IONIC_RXQ_COMP_CSUM_F_IP_BAD	0x20
+#define IONIC_RXQ_COMP_CSUM_F_VLAN	0x40
+#define IONIC_RXQ_COMP_CSUM_F_CALC	0x80
+	u8     pkt_type_color;
+#define IONIC_RXQ_COMP_PKT_TYPE_MASK	0x0f
+};
+
+enum pkt_type {
+	PKT_TYPE_NON_IP     = 0x000,
+	PKT_TYPE_IPV4       = 0x001,
+	PKT_TYPE_IPV4_TCP   = 0x003,
+	PKT_TYPE_IPV4_UDP   = 0x005,
+	PKT_TYPE_IPV6       = 0x008,
+	PKT_TYPE_IPV6_TCP   = 0x018,
+	PKT_TYPE_IPV6_UDP   = 0x028,
+};
+
+enum eth_hw_features {
+	ETH_HW_VLAN_TX_TAG	= BIT(0),
+	ETH_HW_VLAN_RX_STRIP	= BIT(1),
+	ETH_HW_VLAN_RX_FILTER	= BIT(2),
+	ETH_HW_RX_HASH		= BIT(3),
+	ETH_HW_RX_CSUM		= BIT(4),
+	ETH_HW_TX_SG		= BIT(5),
+	ETH_HW_RX_SG		= BIT(6),
+	ETH_HW_TX_CSUM		= BIT(7),
+	ETH_HW_TSO		= BIT(8),
+	ETH_HW_TSO_IPV6		= BIT(9),
+	ETH_HW_TSO_ECN		= BIT(10),
+	ETH_HW_TSO_GRE		= BIT(11),
+	ETH_HW_TSO_GRE_CSUM	= BIT(12),
+	ETH_HW_TSO_IPXIP4	= BIT(13),
+	ETH_HW_TSO_IPXIP6	= BIT(14),
+	ETH_HW_TSO_UDP		= BIT(15),
+	ETH_HW_TSO_UDP_CSUM	= BIT(16),
+};
+
+/**
+ * struct q_control_cmd - Queue control command
+ * @opcode:     opcode
+ * @type:       Queue type
+ * @lif_index:  LIF index
+ * @index:      Queue index
+ * @oper:       Operation (enum q_control_oper)
+ */
+struct q_control_cmd {
+	u8     opcode;
+	u8     type;
+	__le16 lif_index;
+	__le32 index;
+	u8     oper;
+	u8     rsvd[55];
+};
+
+typedef struct admin_comp q_control_comp;
+
+enum q_control_oper {
+	IONIC_Q_DISABLE		= 0,
+	IONIC_Q_ENABLE		= 1,
+	IONIC_Q_HANG_RESET	= 2,
+};
+
+/**
+ * Physical connection type
+ */
+enum phy_type {
+	PHY_TYPE_NONE	= 0,
+	PHY_TYPE_COPPER	= 1,
+	PHY_TYPE_FIBER	= 2,
+};
+
+/**
+ * Transceiver status
+ */
+enum xcvr_state {
+	XCVR_STATE_REMOVED	   = 0,
+	XCVR_STATE_INSERTED	   = 1,
+	XCVR_STATE_PENDING	   = 2,
+	XCVR_STATE_SPROM_READ	   = 3,
+	XCVR_STATE_SPROM_READ_ERR  = 4,
+};
+
+/**
+ * Supported link modes
+ */
+enum xcvr_pid {
+	XCVR_PID_UNKNOWN           = 0,
+
+	/* CU */
+	XCVR_PID_QSFP_100G_CR4     = 1,
+	XCVR_PID_QSFP_40GBASE_CR4  = 2,
+	XCVR_PID_SFP_25GBASE_CR_S  = 3,
+	XCVR_PID_SFP_25GBASE_CR_L  = 4,
+	XCVR_PID_SFP_25GBASE_CR_N  = 5,
+
+	/* Fiber */
+	XCVR_PID_QSFP_100G_AOC    = 50,
+	XCVR_PID_QSFP_100G_ACC    = 51,
+	XCVR_PID_QSFP_100G_SR4    = 52,
+	XCVR_PID_QSFP_100G_LR4    = 53,
+	XCVR_PID_QSFP_100G_ER4    = 54,
+	XCVR_PID_QSFP_40GBASE_ER4 = 55,
+	XCVR_PID_QSFP_40GBASE_SR4 = 56,
+	XCVR_PID_QSFP_40GBASE_LR4 = 57,
+	XCVR_PID_QSFP_40GBASE_AOC = 58,
+	XCVR_PID_SFP_25GBASE_SR   = 59,
+	XCVR_PID_SFP_25GBASE_LR   = 60,
+	XCVR_PID_SFP_25GBASE_ER   = 61,
+	XCVR_PID_SFP_25GBASE_AOC  = 62,
+	XCVR_PID_SFP_10GBASE_SR   = 63,
+	XCVR_PID_SFP_10GBASE_LR   = 64,
+	XCVR_PID_SFP_10GBASE_LRM  = 65,
+	XCVR_PID_SFP_10GBASE_ER   = 66,
+	XCVR_PID_SFP_10GBASE_AOC  = 67,
+	XCVR_PID_SFP_10GBASE_CU   = 68,
+};
+
+/**
+ * Port types
+ */
+enum PortType {
+	PORT_TYPE_NONE = 0,	/* port type not configured */
+	PORT_TYPE_ETH  = 1,	/* port carries ethernet traffic (inband) */
+	PORT_TYPE_MGMT = 2,	/* port carries mgmt traffic (out-of-band) */
+};
+
+/**
+ * Port config state
+ */
+enum PortAdminState {
+	PORT_ADMIN_STATE_NONE = 0,	/* port admin state not configured */
+	PORT_ADMIN_STATE_DOWN = 1,	/* port is admin disabled */
+	PORT_ADMIN_STATE_UP   = 2,	/* port is admin enabled */
+};
+
+/**
+ * Port operational status
+ */
+enum port_oper_status {
+	PORT_OPER_STATUS_NONE  = 0,	/* port is disabled */
+	PORT_OPER_STATUS_UP    = 1,	/* port is linked up */
+	PORT_OPER_STATUS_DOWN  = 2,	/* port link status is down */
+};
+
+/**
+ * Ethernet Forward error correction (fec) modes
+ */
+enum port_fec_type {
+	PORT_FEC_TYPE_NONE = 0,		/* Disabled */
+	PORT_FEC_TYPE_FC   = 1,		/* FireCode */
+	PORT_FEC_TYPE_RS   = 2,		/* ReedSolomon */
+};
+
+/**
+ * Ethernet pause (flow control) modes
+ */
+enum port_pause_type {
+	PORT_PAUSE_TYPE_NONE = 0,	/* Disable Pause */
+	PORT_PAUSE_TYPE_LINK = 1,	/* Link level pause */
+	PORT_PAUSE_TYPE_PFC  = 2,	/* Priority-Flow control */
+};
+
+/**
+ * Loopback modes
+ */
+enum port_loopback_mode {
+	PORT_LOOPBACK_MODE_NONE = 0,	/* Disable loopback */
+	PORT_LOOPBACK_MODE_MAC  = 1,	/* MAC loopback */
+	PORT_LOOPBACK_MODE_PHY  = 2,	/* PHY/Serdes loopback */
+};
+
+/**
+ * Transceiver Status information
+ * @state:    Transceiver status (enum xcvr_state)
+ * @phy:      Physical connection type (enum phy_type)
+ * @pid:      Transceiver link mode (enum pid)
+ * @sprom:    Transceiver sprom contents
+ */
+struct xcvr_status {
+	u8     state;
+	u8     phy;
+	__le16 pid;
+	u8     sprom[256];
+};
+
+/**
+ * Port configuration
+ * @speed:              port speed (in Mbps)
+ * @mtu:                mtu
+ * @state:              port admin state (enum port_admin_state)
+ * @an_enable:          autoneg enable
+ * @fec_type:           fec type (enum port_fec_type)
+ * @pause_type:         pause type (enum port_pause_type)
+ * @loopback_mode:      loopback mode (enum port_loopback_mode)
+ */
+union port_config {
+	struct {
+#define IONIC_SPEED_100G	100000	/* 100G in Mbps */
+#define IONIC_SPEED_50G		50000	/* 50G in Mbps */
+#define IONIC_SPEED_40G		40000	/* 40G in Mbps */
+#define IONIC_SPEED_25G		25000	/* 25G in Mbps */
+#define IONIC_SPEED_10G		10000	/* 10G in Mbps */
+#define IONIC_SPEED_1G		1000	/* 1G in Mbps */
+		__le32 speed;
+		__le32 mtu;
+		u8     state;
+		u8     an_enable;
+		u8     fec_type;
+#define IONIC_PAUSE_TYPE_MASK		0x0f
+#define IONIC_PAUSE_FLAGS_MASK		0xf0
+#define IONIC_PAUSE_F_TX		0x10
+#define IONIC_PAUSE_F_RX		0x20
+		u8     pause_type;
+		u8     loopback_mode;
+	};
+	__le32 words[64];
+};
+
+/**
+ * Port Status information
+ * @status:             link status (enum port_oper_status)
+ * @id:                 port id
+ * @speed:              link speed (in Mbps)
+ * @xcvr:               tranceiver status
+ */
+struct port_status {
+	__le32 id;
+	__le32 speed;
+	u8     status;
+	u8     rsvd[51];
+	struct xcvr_status  xcvr;
+};
+
+/**
+ * struct port_identify_cmd - Port identify command
+ * @opcode:     opcode
+ * @index:      port index
+ * @ver:        Highest version of identify supported by driver
+ */
+struct port_identify_cmd {
+	u8 opcode;
+	u8 index;
+	u8 ver;
+	u8 rsvd[61];
+};
+
+/**
+ * struct port_identify_comp - Port identify command completion
+ * @status: The status of the command (enum status_code)
+ * @ver:    Version of identify returned by device
+ */
+struct port_identify_comp {
+	u8 status;
+	u8 ver;
+	u8 rsvd[14];
+};
+
+/**
+ * struct port_init_cmd - Port initialization command
+ * @opcode:     opcode
+ * @index:      port index
+ * @info_pa:    destination address for port info (struct port_info)
+ */
+struct port_init_cmd {
+	u8     opcode;
+	u8     index;
+	u8     rsvd[6];
+	__le64 info_pa;
+	u8     rsvd2[48];
+};
+
+/**
+ * struct port_init_comp - Port initialization command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct port_init_comp {
+	u8 status;
+	u8 rsvd[15];
+};
+
+/**
+ * struct port_reset_cmd - Port reset command
+ * @opcode:     opcode
+ * @index:      port index
+ */
+struct port_reset_cmd {
+	u8 opcode;
+	u8 index;
+	u8 rsvd[62];
+};
+
+/**
+ * struct port_reset_comp - Port reset command completion
+ * @status: The status of the command (enum status_code)
+ */
+struct port_reset_comp {
+	u8 status;
+	u8 rsvd[15];
+};
+
+/**
+ * enum ionic_port_attr - List of device attributes
+ */
+enum ionic_port_attr {
+	IONIC_PORT_ATTR_STATE		= 0,
+	IONIC_PORT_ATTR_SPEED		= 1,
+	IONIC_PORT_ATTR_MTU		= 2,
+	IONIC_PORT_ATTR_AUTONEG		= 3,
+	IONIC_PORT_ATTR_FEC		= 4,
+	IONIC_PORT_ATTR_PAUSE		= 5,
+	IONIC_PORT_ATTR_LOOPBACK	= 6,
+};
+
+/**
+ * struct port_setattr_cmd - Set port attributes on the NIC
+ * @opcode:     Opcode
+ * @index:      port index
+ * @attr:       Attribute type (enum ionic_port_attr)
+ */
+struct port_setattr_cmd {
+	u8     opcode;
+	u8     index;
+	u8     attr;
+	u8     rsvd;
+	union {
+		u8      state;
+		__le32  speed;
+		__le32  mtu;
+		u8      an_enable;
+		u8      fec_type;
+		u8      pause_type;
+		u8      loopback_mode;
+		u8      rsvd2[60];
+	};
+};
+
+/**
+ * struct port_setattr_comp - Port set attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @color:      Color bit
+ */
+struct port_setattr_comp {
+	u8     status;
+	u8     rsvd[14];
+	u8     color;
+};
+
+/**
+ * struct port_getattr_cmd - Get port attributes from the NIC
+ * @opcode:     Opcode
+ * @index:      port index
+ * @attr:       Attribute type (enum ionic_port_attr)
+ */
+struct port_getattr_cmd {
+	u8     opcode;
+	u8     index;
+	u8     attr;
+	u8     rsvd[61];
+};
+
+/**
+ * struct port_getattr_comp - Port get attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @color:      Color bit
+ */
+struct port_getattr_comp {
+	u8     status;
+	u8     rsvd[3];
+	union {
+		u8      state;
+		__le32  speed;
+		__le32  mtu;
+		u8      an_enable;
+		u8      fec_type;
+		u8      pause_type;
+		u8      loopback_mode;
+		u8      rsvd2[11];
+	};
+	u8     color;
+};
+
+/**
+ * struct lif_status - Lif status register
+ * @eid:             most recent NotifyQ event id
+ * @port_num:        port the lif is connected to
+ * @link_status:     port status (enum port_oper_status)
+ * @link_speed:      speed of link in Mbps
+ * @link_down_count: number of times link status changes
+ */
+struct lif_status {
+	__le64 eid;
+	u8     port_num;
+	u8     rsvd;
+	__le16 link_status;
+	__le32 link_speed;		/* units of 1Mbps: eg 10000 = 10Gbps */
+	__le16 link_down_count;
+	u8      rsvd2[46];
+};
+
+/**
+ * struct lif_reset_cmd - LIF reset command
+ * @opcode:    opcode
+ * @index:     LIF index
+ */
+struct lif_reset_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 index;
+	__le32 rsvd2[15];
+};
+
+typedef struct admin_comp lif_reset_comp;
+
+enum dev_state {
+	IONIC_DEV_DISABLE	= 0,
+	IONIC_DEV_ENABLE	= 1,
+	IONIC_DEV_HANG_RESET	= 2,
+};
+
+/**
+ * enum dev_attr - List of device attributes
+ */
+enum dev_attr {
+	IONIC_DEV_ATTR_STATE    = 0,
+	IONIC_DEV_ATTR_NAME     = 1,
+	IONIC_DEV_ATTR_FEATURES = 2,
+};
+
+/**
+ * struct dev_setattr_cmd - Set Device attributes on the NIC
+ * @opcode:     Opcode
+ * @attr:       Attribute type (enum dev_attr)
+ * @state:      Device state (enum dev_state)
+ * @name:       The bus info, e.g. PCI slot-device-function, 0 terminated
+ * @features:   Device features
+ */
+struct dev_setattr_cmd {
+	u8     opcode;
+	u8     attr;
+	__le16 rsvd;
+	union {
+		u8      state;
+		char    name[IONIC_IFNAMSIZ];
+		__le64  features;
+		u8      rsvd2[60];
+	};
+};
+
+/**
+ * struct dev_setattr_comp - Device set attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @features:   Device features
+ * @color:      Color bit
+ */
+struct dev_setattr_comp {
+	u8     status;
+	u8     rsvd[3];
+	union {
+		__le64  features;
+		u8      rsvd2[11];
+	};
+	u8     color;
+};
+
+/**
+ * struct dev_getattr_cmd - Get Device attributes from the NIC
+ * @opcode:     opcode
+ * @attr:       Attribute type (enum dev_attr)
+ */
+struct dev_getattr_cmd {
+	u8     opcode;
+	u8     attr;
+	u8     rsvd[62];
+};
+
+/**
+ * struct dev_setattr_comp - Device set attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @features:   Device features
+ * @color:      Color bit
+ */
+struct dev_getattr_comp {
+	u8     status;
+	u8     rsvd[3];
+	union {
+		__le64  features;
+		u8      rsvd2[11];
+	};
+	u8     color;
+};
+
+/**
+ * RSS parameters
+ */
+#define IONIC_RSS_HASH_KEY_SIZE		40
+
+enum rss_hash_types {
+	IONIC_RSS_TYPE_IPV4	= BIT(0),
+	IONIC_RSS_TYPE_IPV4_TCP	= BIT(1),
+	IONIC_RSS_TYPE_IPV4_UDP	= BIT(2),
+	IONIC_RSS_TYPE_IPV6	= BIT(3),
+	IONIC_RSS_TYPE_IPV6_TCP	= BIT(4),
+	IONIC_RSS_TYPE_IPV6_UDP	= BIT(5),
+};
+
+/**
+ * enum lif_attr - List of LIF attributes
+ */
+enum lif_attr {
+	IONIC_LIF_ATTR_STATE        = 0,
+	IONIC_LIF_ATTR_NAME         = 1,
+	IONIC_LIF_ATTR_MTU          = 2,
+	IONIC_LIF_ATTR_MAC          = 3,
+	IONIC_LIF_ATTR_FEATURES     = 4,
+	IONIC_LIF_ATTR_RSS          = 5,
+};
+
+/**
+ * struct lif_setattr_cmd - Set LIF attributes on the NIC
+ * @opcode:     Opcode
+ * @type:       Attribute type (enum lif_attr)
+ * @index:      LIF index
+ * @state:      lif state (enum lif_state)
+ * @name:       The netdev name string, 0 terminated
+ * @mtu:        Mtu
+ * @mac:        Station mac
+ * @features:   Features (enum eth_hw_features)
+ * @rss:        RSS properties
+ *              @types:     The hash types to enable (see rss_hash_types).
+ *              @key:       The hash secret key.
+ *              @addr:      Address for the indirection table shared memory.
+ */
+struct lif_setattr_cmd {
+	u8     opcode;
+	u8     attr;
+	__le16 index;
+	union {
+		u8      state;
+		char    name[IONIC_IFNAMSIZ];
+		__le32  mtu;
+		u8      mac[6];
+		__le64  features;
+		struct {
+			__le16 types;
+			u8     key[IONIC_RSS_HASH_KEY_SIZE];
+			u8     rsvd[6];
+			__le64 addr;
+		} rss;
+		u8      rsvd[60];
+	};
+};
+
+/**
+ * struct lif_setattr_comp - LIF set attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @comp_index: The index in the descriptor ring for which this
+ *              is the completion.
+ * @features:   features (enum eth_hw_features)
+ * @color:      Color bit
+ */
+struct lif_setattr_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	union {
+		__le64  features;
+		u8      rsvd2[11];
+	};
+	u8     color;
+};
+
+/**
+ * struct lif_getattr_cmd - Get LIF attributes from the NIC
+ * @opcode:     Opcode
+ * @attr:       Attribute type (enum lif_attr)
+ * @index:      LIF index
+ */
+struct lif_getattr_cmd {
+	u8     opcode;
+	u8     attr;
+	__le16 index;
+	u8     rsvd[60];
+};
+
+/**
+ * struct lif_getattr_comp - LIF get attr command completion
+ * @status:     The status of the command (enum status_code)
+ * @comp_index: The index in the descriptor ring for which this
+ *              is the completion.
+ * @state:      lif state (enum lif_state)
+ * @name:       The netdev name string, 0 terminated
+ * @mtu:        Mtu
+ * @mac:        Station mac
+ * @features:   Features (enum eth_hw_features)
+ * @color:      Color bit
+ */
+struct lif_getattr_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	union {
+		u8      state;
+		__le32  mtu;
+		u8      mac[6];
+		__le64  features;
+		u8      rsvd2[11];
+	};
+	u8     color;
+};
+
+enum rx_mode {
+	RX_MODE_F_UNICAST	= BIT(0),
+	RX_MODE_F_MULTICAST	= BIT(1),
+	RX_MODE_F_BROADCAST	= BIT(2),
+	RX_MODE_F_PROMISC	= BIT(3),
+	RX_MODE_F_ALLMULTI	= BIT(4),
+};
+
+/**
+ * struct rx_mode_set_cmd - Set LIF's Rx mode command
+ * @opcode:     opcode
+ * @lif_index:  LIF index
+ * @rx_mode:    Rx mode flags:
+ *                  RX_MODE_F_UNICAST: Accept known unicast packets.
+ *                  RX_MODE_F_MULTICAST: Accept known multicast packets.
+ *                  RX_MODE_F_BROADCAST: Accept broadcast packets.
+ *                  RX_MODE_F_PROMISC: Accept any packets.
+ *                  RX_MODE_F_ALLMULTI: Accept any multicast packets.
+ */
+struct rx_mode_set_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	__le16 rx_mode;
+	__le16 rsvd2[29];
+};
+
+typedef struct admin_comp rx_mode_set_comp;
+
+enum rx_filter_match_type {
+	RX_FILTER_MATCH_VLAN = 0,
+	RX_FILTER_MATCH_MAC,
+	RX_FILTER_MATCH_MAC_VLAN,
+};
+
+/**
+ * struct rx_filter_add_cmd - Add LIF Rx filter command
+ * @opcode:     opcode
+ * @qtype:      Queue type
+ * @lif_index:  LIF index
+ * @qid:        Queue ID
+ * @match:      Rx filter match type.  (See RX_FILTER_MATCH_xxx)
+ * @vlan:       VLAN ID
+ * @addr:       MAC address (network-byte order)
+ */
+struct rx_filter_add_cmd {
+	u8     opcode;
+	u8     qtype;
+	__le16 lif_index;
+	__le32 qid;
+	__le16 match;
+	union {
+		struct {
+			__le16 vlan;
+		} vlan;
+		struct {
+			u8     addr[6];
+		} mac;
+		struct {
+			__le16 vlan;
+			u8     addr[6];
+		} mac_vlan;
+		u8 rsvd[54];
+	};
+};
+
+/**
+ * struct rx_filter_add_comp - Add LIF Rx filter command completion
+ * @status:     The status of the command (enum status_code)
+ * @comp_index: The index in the descriptor ring for which this
+ *              is the completion.
+ * @filter_id:  Filter ID
+ * @color:      Color bit.
+ */
+struct rx_filter_add_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	__le32 filter_id;
+	u8     rsvd2[7];
+	u8     color;
+};
+
+/**
+ * struct rx_filter_del_cmd - Delete LIF Rx filter command
+ * @opcode:     opcode
+ * @lif_index:  LIF index
+ * @filter_id:  Filter ID
+ */
+struct rx_filter_del_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	__le32 filter_id;
+	u8     rsvd2[56];
+};
+
+typedef struct admin_comp rx_filter_del_comp;
+
+/**
+ * struct qos_identify_cmd - QoS identify command
+ * @opcode:    opcode
+ * @ver:     Highest version of identify supported by driver
+ *
+ */
+struct qos_identify_cmd {
+	u8 opcode;
+	u8 ver;
+	u8 rsvd[62];
+};
+
+/**
+ * struct qos_identify_comp - QoS identify command completion
+ * @status: The status of the command (enum status_code)
+ * @ver:    Version of identify returned by device
+ */
+struct qos_identify_comp {
+	u8 status;
+	u8 ver;
+	u8 rsvd[14];
+};
+
+#define IONIC_QOS_CLASS_MAX		7
+#define IONIC_QOS_CLASS_NAME_SZ		32
+#define IONIC_QOS_DSCP_MAX_VALUES	64
+
+/**
+ * enum qos_class
+ */
+enum qos_class {
+	QOS_CLASS_DEFAULT		= 0,
+	QOS_CLASS_USER_DEFINED_1	= 1,
+	QOS_CLASS_USER_DEFINED_2	= 2,
+	QOS_CLASS_USER_DEFINED_3	= 3,
+	QOS_CLASS_USER_DEFINED_4	= 4,
+	QOS_CLASS_USER_DEFINED_5	= 5,
+	QOS_CLASS_USER_DEFINED_6	= 6,
+};
+
+/**
+ * enum qos_class_type - Traffic classification criteria
+ */
+enum qos_class_type {
+	QOS_CLASS_TYPE_NONE	= 0,
+	QOS_CLASS_TYPE_PCP	= 1,	/* Dot1Q pcp */
+	QOS_CLASS_TYPE_DSCP	= 2,	/* IP dscp */
+};
+
+/**
+ * enum qos_sched_type - Qos class scheduling type
+ */
+enum qos_sched_type {
+	QOS_SCHED_TYPE_STRICT	= 0,	/* Strict priority */
+	QOS_SCHED_TYPE_DWRR	= 1,	/* Deficit weighted round-robin */
+};
+
+/**
+ * union qos_config - Qos configuration structure
+ * @flags:		Configuration flags
+ *	IONIC_QOS_CONFIG_F_ENABLE		enable
+ *	IONIC_QOS_CONFIG_F_DROP			drop/nodrop
+ *	IONIC_QOS_CONFIG_F_RW_DOT1Q_PCP		enable dot1q pcp rewrite
+ *	IONIC_QOS_CONFIG_F_RW_IP_DSCP		enable ip dscp rewrite
+ * @sched_type:		Qos class scheduling type (enum qos_sched_type)
+ * @class_type:		Qos class type (enum qos_class_type)
+ * @pause_type:		Qos pause type (enum qos_pause_type)
+ * @name:		Qos class name
+ * @mtu:		MTU of the class
+ * @pfc_dot1q_pcp:	Pcp value for pause frames (valid iff F_NODROP)
+ * @dwrr_weight:	Qos class scheduling weight
+ * @strict_rlmt:	Rate limit for strict priority scheduling
+ * @rw_dot1q_pcp:	Rewrite dot1q pcp to this value	(valid iff F_RW_DOT1Q_PCP)
+ * @rw_ip_dscp:		Rewrite ip dscp to this value	(valid iff F_RW_IP_DSCP)
+ * @dot1q_pcp:		Dot1q pcp value
+ * @ndscp:		Number of valid dscp values in the ip_dscp field
+ * @ip_dscp:		IP dscp values
+ */
+union qos_config {
+	struct {
+#define IONIC_QOS_CONFIG_F_ENABLE		BIT(0)
+#define IONIC_QOS_CONFIG_F_DROP			BIT(1)
+#define IONIC_QOS_CONFIG_F_RW_DOT1Q_PCP		BIT(2)
+#define IONIC_QOS_CONFIG_F_RW_IP_DSCP		BIT(3)
+		u8      flags;
+		u8      sched_type;
+		u8      class_type;
+		u8      pause_type;
+		char    name[IONIC_QOS_CLASS_NAME_SZ];
+		__le32  mtu;
+		/* flow control */
+		u8      pfc_cos;
+		/* scheduler */
+		union {
+			u8      dwrr_weight;
+			__le64  strict_rlmt;
+		};
+		/* marking */
+		union {
+			u8      rw_dot1q_pcp;
+			u8      rw_ip_dscp;
+		};
+		/* classification */
+		union {
+			u8      dot1q_pcp;
+			struct {
+				u8      ndscp;
+				u8      ip_dscp[IONIC_QOS_DSCP_MAX_VALUES];
+			};
+		};
+	};
+	__le32  words[64];
+};
+
+/**
+ * union qos_identity - QoS identity structure
+ * @version:	Version of the identify structure
+ * @type:	QoS system type
+ * @nclasses:	Number of usable QoS classes
+ * @config:	Current configuration of classes
+ */
+union qos_identity {
+	struct {
+		u8     version;
+		u8     type;
+		u8     rsvd[62];
+		union  qos_config config[IONIC_QOS_CLASS_MAX];
+	};
+	__le32 words[512];
+};
+
+/**
+ * struct qos_init_cmd - QoS config init command
+ * @opcode:	Opcode
+ * @group:	Qos class id
+ * @info_pa:	destination address for qos info
+ */
+struct qos_init_cmd {
+	u8     opcode;
+	u8     group;
+	u8     rsvd[6];
+	__le64 info_pa;
+	u8     rsvd1[48];
+};
+
+typedef struct admin_comp qos_init_comp;
+
+/**
+ * struct qos_reset_cmd - Qos config reset command
+ * @opcode:	Opcode
+ */
+struct qos_reset_cmd {
+	u8    opcode;
+	u8    group;
+	u8    rsvd[62];
+};
+
+typedef struct admin_comp qos_reset_comp;
+
+/**
+ * struct fw_download_cmd - Firmware download command
+ * @opcode:	opcode
+ * @addr:	dma address of the firmware buffer
+ * @offset:	offset of the firmware buffer within the full image
+ * @length:	number of valid bytes in the firmware buffer
+ */
+struct fw_download_cmd {
+	u8     opcode;
+	u8     rsvd[3];
+	__le32 offset;
+	__le64 addr;
+	__le32 length;
+};
+
+typedef struct admin_comp fw_download_comp;
+
+enum fw_control_oper {
+	IONIC_FW_RESET		= 0,	/* Reset firmware */
+	IONIC_FW_INSTALL	= 1,	/* Install firmware */
+	IONIC_FW_ACTIVATE	= 2,	/* Activate firmware */
+};
+
+/**
+ * struct fw_control_cmd - Firmware control command
+ * @opcode:    opcode
+ * @oper:      firmware control operation (enum fw_control_oper)
+ * @slot:      slot to activate
+ */
+struct fw_control_cmd {
+	u8  opcode;
+	u8  rsvd[3];
+	u8  oper;
+	u8  slot;
+	u8  rsvd1[58];
+};
+
+/**
+ * struct fw_control_comp - Firmware control copletion
+ * @opcode:    opcode
+ * @slot:      slot where the firmware was installed
+ */
+struct fw_control_comp {
+	u8     status;
+	u8     rsvd;
+	__le16 comp_index;
+	u8     slot;
+	u8     rsvd1[10];
+	u8     color;
+};
+
+/******************************************************************
+ ******************* RDMA Commands ********************************
+ ******************************************************************/
+
+/**
+ * struct rdma_reset_cmd - Reset RDMA LIF cmd
+ * @opcode:        opcode
+ * @lif_index:     lif index
+ *
+ * There is no rdma specific dev command completion struct.  Completion uses
+ * the common struct admin_comp.  Only the status is indicated.  Nonzero status
+ * means the LIF does not support rdma.
+ **/
+struct rdma_reset_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	u8     rsvd2[60];
+};
+
+/**
+ * struct rdma_queue_cmd - Create RDMA Queue command
+ * @opcode:        opcode, 52, 53
+ * @lif_index      lif index
+ * @qid_ver:       (qid | (rdma version << 24))
+ * @cid:           intr, eq_id, or cq_id
+ * @dbid:          doorbell page id
+ * @depth_log2:    log base two of queue depth
+ * @stride_log2:   log base two of queue stride
+ * @dma_addr:      address of the queue memory
+ * @xxx_table_index: temporary, but should not need pgtbl for contig. queues.
+ *
+ * The same command struct is used to create an rdma event queue, completion
+ * queue, or rdma admin queue.  The cid is an interrupt number for an event
+ * queue, an event queue id for a completion queue, or a completion queue id
+ * for an rdma admin queue.
+ *
+ * The queue created via a dev command must be contiguous in dma space.
+ *
+ * The dev commands are intended only to be used during driver initialization,
+ * to create queues supporting the rdma admin queue.  Other queues, and other
+ * types of rdma resources like memory regions, will be created and registered
+ * via the rdma admin queue, and will support a more complete interface
+ * providing scatter gather lists for larger, scattered queue buffers and
+ * memory registration.
+ *
+ * There is no rdma specific dev command completion struct.  Completion uses
+ * the common struct admin_comp.  Only the status is indicated.
+ **/
+struct rdma_queue_cmd {
+	u8     opcode;
+	u8     rsvd;
+	__le16 lif_index;
+	__le32 qid_ver;
+	__le32 cid;
+	__le16 dbid;
+	u8     depth_log2;
+	u8     stride_log2;
+	__le64 dma_addr;
+	u8     rsvd2[36];
+	__le32 xxx_table_index;
+};
+
+/******************************************************************
+ ******************* Notify Events ********************************
+ ******************************************************************/
+
+/**
+ * struct notifyq_event
+ * @eid:   event number
+ * @ecode: event code
+ * @data:  unspecified data about the event
+ *
+ * This is the generic event report struct from which the other
+ * actual events will be formed.
+ */
+struct notifyq_event {
+	__le64 eid;
+	__le16 ecode;
+	u8     data[54];
+};
+
+/**
+ * struct link_change_event
+ * @eid:		event number
+ * @ecode:		event code = EVENT_OPCODE_LINK_CHANGE
+ * @link_status:	link up or down, with error bits (enum port_status)
+ * @link_speed:		speed of the network link
+ *
+ * Sent when the network link state changes between UP and DOWN
+ */
+struct link_change_event {
+	__le64 eid;
+	__le16 ecode;
+	__le16 link_status;
+	__le32 link_speed;	/* units of 1Mbps: e.g. 10000 = 10Gbps */
+	u8     rsvd[48];
+};
+
+/**
+ * struct reset_event
+ * @eid:		event number
+ * @ecode:		event code = EVENT_OPCODE_RESET
+ * @reset_code:		reset type
+ * @state:		0=pending, 1=complete, 2=error
+ *
+ * Sent when the NIC or some subsystem is going to be or
+ * has been reset.
+ */
+struct reset_event {
+	__le64 eid;
+	__le16 ecode;
+	u8     reset_code;
+	u8     state;
+	u8     rsvd[52];
+};
+
+/**
+ * struct heartbeat_event
+ * @eid:	event number
+ * @ecode:	event code = EVENT_OPCODE_HEARTBEAT
+ *
+ * Sent periodically by the NIC to indicate continued health
+ */
+struct heartbeat_event {
+	__le64 eid;
+	__le16 ecode;
+	u8     rsvd[54];
+};
+
+/**
+ * struct log_event
+ * @eid:	event number
+ * @ecode:	event code = EVENT_OPCODE_LOG
+ * @data:	log data
+ *
+ * Sent to notify the driver of an internal error.
+ */
+struct log_event {
+	__le64 eid;
+	__le16 ecode;
+	u8     data[54];
+};
+
+/**
+ * struct port_stats
+ */
+struct port_stats {
+	__le64 frames_rx_ok;
+	__le64 frames_rx_all;
+	__le64 frames_rx_bad_fcs;
+	__le64 frames_rx_bad_all;
+	__le64 octets_rx_ok;
+	__le64 octets_rx_all;
+	__le64 frames_rx_unicast;
+	__le64 frames_rx_multicast;
+	__le64 frames_rx_broadcast;
+	__le64 frames_rx_pause;
+	__le64 frames_rx_bad_length;
+	__le64 frames_rx_undersized;
+	__le64 frames_rx_oversized;
+	__le64 frames_rx_fragments;
+	__le64 frames_rx_jabber;
+	__le64 frames_rx_pripause;
+	__le64 frames_rx_stomped_crc;
+	__le64 frames_rx_too_long;
+	__le64 frames_rx_vlan_good;
+	__le64 frames_rx_dropped;
+	__le64 frames_rx_less_than_64b;
+	__le64 frames_rx_64b;
+	__le64 frames_rx_65b_127b;
+	__le64 frames_rx_128b_255b;
+	__le64 frames_rx_256b_511b;
+	__le64 frames_rx_512b_1023b;
+	__le64 frames_rx_1024b_1518b;
+	__le64 frames_rx_1519b_2047b;
+	__le64 frames_rx_2048b_4095b;
+	__le64 frames_rx_4096b_8191b;
+	__le64 frames_rx_8192b_9215b;
+	__le64 frames_rx_other;
+	__le64 frames_tx_ok;
+	__le64 frames_tx_all;
+	__le64 frames_tx_bad;
+	__le64 octets_tx_ok;
+	__le64 octets_tx_total;
+	__le64 frames_tx_unicast;
+	__le64 frames_tx_multicast;
+	__le64 frames_tx_broadcast;
+	__le64 frames_tx_pause;
+	__le64 frames_tx_pripause;
+	__le64 frames_tx_vlan;
+	__le64 frames_tx_less_than_64b;
+	__le64 frames_tx_64b;
+	__le64 frames_tx_65b_127b;
+	__le64 frames_tx_128b_255b;
+	__le64 frames_tx_256b_511b;
+	__le64 frames_tx_512b_1023b;
+	__le64 frames_tx_1024b_1518b;
+	__le64 frames_tx_1519b_2047b;
+	__le64 frames_tx_2048b_4095b;
+	__le64 frames_tx_4096b_8191b;
+	__le64 frames_tx_8192b_9215b;
+	__le64 frames_tx_other;
+	__le64 frames_tx_pri_0;
+	__le64 frames_tx_pri_1;
+	__le64 frames_tx_pri_2;
+	__le64 frames_tx_pri_3;
+	__le64 frames_tx_pri_4;
+	__le64 frames_tx_pri_5;
+	__le64 frames_tx_pri_6;
+	__le64 frames_tx_pri_7;
+	__le64 frames_rx_pri_0;
+	__le64 frames_rx_pri_1;
+	__le64 frames_rx_pri_2;
+	__le64 frames_rx_pri_3;
+	__le64 frames_rx_pri_4;
+	__le64 frames_rx_pri_5;
+	__le64 frames_rx_pri_6;
+	__le64 frames_rx_pri_7;
+	__le64 tx_pripause_0_1us_count;
+	__le64 tx_pripause_1_1us_count;
+	__le64 tx_pripause_2_1us_count;
+	__le64 tx_pripause_3_1us_count;
+	__le64 tx_pripause_4_1us_count;
+	__le64 tx_pripause_5_1us_count;
+	__le64 tx_pripause_6_1us_count;
+	__le64 tx_pripause_7_1us_count;
+	__le64 rx_pripause_0_1us_count;
+	__le64 rx_pripause_1_1us_count;
+	__le64 rx_pripause_2_1us_count;
+	__le64 rx_pripause_3_1us_count;
+	__le64 rx_pripause_4_1us_count;
+	__le64 rx_pripause_5_1us_count;
+	__le64 rx_pripause_6_1us_count;
+	__le64 rx_pripause_7_1us_count;
+	__le64 rx_pause_1us_count;
+	__le64 frames_tx_truncated;
+};
+
+struct mgmt_port_stats {
+	__le64 frames_rx_ok;
+	__le64 frames_rx_all;
+	__le64 frames_rx_bad_fcs;
+	__le64 frames_rx_bad_all;
+	__le64 octets_rx_ok;
+	__le64 octets_rx_all;
+	__le64 frames_rx_unicast;
+	__le64 frames_rx_multicast;
+	__le64 frames_rx_broadcast;
+	__le64 frames_rx_pause;
+	__le64 frames_rx_bad_length0;
+	__le64 frames_rx_undersized1;
+	__le64 frames_rx_oversized2;
+	__le64 frames_rx_fragments3;
+	__le64 frames_rx_jabber4;
+	__le64 frames_rx_64b5;
+	__le64 frames_rx_65b_127b6;
+	__le64 frames_rx_128b_255b7;
+	__le64 frames_rx_256b_511b8;
+	__le64 frames_rx_512b_1023b9;
+	__le64 frames_rx_1024b_1518b0;
+	__le64 frames_rx_gt_1518b1;
+	__le64 frames_rx_fifo_full2;
+	__le64 frames_tx_ok3;
+	__le64 frames_tx_all4;
+	__le64 frames_tx_bad5;
+	__le64 octets_tx_ok6;
+	__le64 octets_tx_total7;
+	__le64 frames_tx_unicast8;
+	__le64 frames_tx_multicast9;
+	__le64 frames_tx_broadcast0;
+	__le64 frames_tx_pause1;
+};
+
+/**
+ * struct port_identity - port identity structure
+ * @version:        identity structure version
+ * @type:           type of port (enum port_type)
+ * @num_lanes:      number of lanes for the port
+ * @autoneg:        autoneg supported
+ * @min_frame_size: minimum frame size supported
+ * @max_frame_size: maximum frame size supported
+ * @fec_type:       supported fec types
+ * @pause_type:     supported pause types
+ * @loopback_mode:  supported loopback mode
+ * @speeds:         supported speeds
+ * @config:         current port configuration
+ */
+union port_identity {
+	struct {
+		u8     version;
+		u8     type;
+		u8     num_lanes;
+		u8     autoneg;
+		__le32 min_frame_size;
+		__le32 max_frame_size;
+		u8     fec_type[4];
+		u8     pause_type[2];
+		u8     loopback_mode[2];
+		__le32 speeds[16];
+		u8     rsvd2[44];
+		union port_config config;
+	};
+	__le32 words[512];
+};
+
+/**
+ * struct port_info - port info structure
+ * @port_status:     port status
+ * @port_stats:      port stats
+ */
+struct port_info {
+	union port_config config;
+	struct port_status status;
+	struct port_stats stats;
+};
+
+/**
+ * struct lif_stats
+ */
+struct lif_stats {
+	/* RX */
+	__le64 rx_ucast_bytes;
+	__le64 rx_ucast_packets;
+	__le64 rx_mcast_bytes;
+	__le64 rx_mcast_packets;
+	__le64 rx_bcast_bytes;
+	__le64 rx_bcast_packets;
+	__le64 rsvd0;
+	__le64 rsvd1;
+	/* RX drops */
+	__le64 rx_ucast_drop_bytes;
+	__le64 rx_ucast_drop_packets;
+	__le64 rx_mcast_drop_bytes;
+	__le64 rx_mcast_drop_packets;
+	__le64 rx_bcast_drop_bytes;
+	__le64 rx_bcast_drop_packets;
+	__le64 rx_dma_error;
+	__le64 rsvd2;
+	/* TX */
+	__le64 tx_ucast_bytes;
+	__le64 tx_ucast_packets;
+	__le64 tx_mcast_bytes;
+	__le64 tx_mcast_packets;
+	__le64 tx_bcast_bytes;
+	__le64 tx_bcast_packets;
+	__le64 rsvd3;
+	__le64 rsvd4;
+	/* TX drops */
+	__le64 tx_ucast_drop_bytes;
+	__le64 tx_ucast_drop_packets;
+	__le64 tx_mcast_drop_bytes;
+	__le64 tx_mcast_drop_packets;
+	__le64 tx_bcast_drop_bytes;
+	__le64 tx_bcast_drop_packets;
+	__le64 tx_dma_error;
+	__le64 rsvd5;
+	/* Rx Queue/Ring drops */
+	__le64 rx_queue_disabled;
+	__le64 rx_queue_empty;
+	__le64 rx_queue_error;
+	__le64 rx_desc_fetch_error;
+	__le64 rx_desc_data_error;
+	__le64 rsvd6;
+	__le64 rsvd7;
+	__le64 rsvd8;
+	/* Tx Queue/Ring drops */
+	__le64 tx_queue_disabled;
+	__le64 tx_queue_error;
+	__le64 tx_desc_fetch_error;
+	__le64 tx_desc_data_error;
+	__le64 rsvd9;
+	__le64 rsvd10;
+	__le64 rsvd11;
+	__le64 rsvd12;
+
+	/* RDMA/ROCE TX */
+	__le64 tx_rdma_ucast_bytes;
+	__le64 tx_rdma_ucast_packets;
+	__le64 tx_rdma_mcast_bytes;
+	__le64 tx_rdma_mcast_packets;
+	__le64 tx_rdma_cnp_packets;
+	__le64 rsvd13;
+	__le64 rsvd14;
+	__le64 rsvd15;
+
+	/* RDMA/ROCE RX */
+	__le64 rx_rdma_ucast_bytes;
+	__le64 rx_rdma_ucast_packets;
+	__le64 rx_rdma_mcast_bytes;
+	__le64 rx_rdma_mcast_packets;
+	__le64 rx_rdma_cnp_packets;
+	__le64 rx_rdma_ecn_packets;
+	__le64 rsvd16;
+	__le64 rsvd17;
+
+	__le64 rsvd18;
+	__le64 rsvd19;
+	__le64 rsvd20;
+	__le64 rsvd21;
+	__le64 rsvd22;
+	__le64 rsvd23;
+	__le64 rsvd24;
+	__le64 rsvd25;
+
+	__le64 rsvd26;
+	__le64 rsvd27;
+	__le64 rsvd28;
+	__le64 rsvd29;
+	__le64 rsvd30;
+	__le64 rsvd31;
+	__le64 rsvd32;
+	__le64 rsvd33;
+
+	__le64 rsvd34;
+	__le64 rsvd35;
+	__le64 rsvd36;
+	__le64 rsvd37;
+	__le64 rsvd38;
+	__le64 rsvd39;
+	__le64 rsvd40;
+	__le64 rsvd41;
+
+	__le64 rsvd42;
+	__le64 rsvd43;
+	__le64 rsvd44;
+	__le64 rsvd45;
+	__le64 rsvd46;
+	__le64 rsvd47;
+	__le64 rsvd48;
+	__le64 rsvd49;
+
+	/* RDMA/ROCE REQ Error/Debugs (768 - 895) */
+	__le64 rdma_req_rx_pkt_seq_err;
+	__le64 rdma_req_rx_rnr_retry_err;
+	__le64 rdma_req_rx_remote_access_err;
+	__le64 rdma_req_rx_remote_inv_req_err;
+	__le64 rdma_req_rx_remote_oper_err;
+	__le64 rdma_req_rx_implied_nak_seq_err;
+	__le64 rdma_req_rx_cqe_err;
+	__le64 rdma_req_rx_cqe_flush_err;
+
+	__le64 rdma_req_rx_dup_responses;
+	__le64 rdma_req_rx_invalid_packets;
+	__le64 rdma_req_tx_local_access_err;
+	__le64 rdma_req_tx_local_oper_err;
+	__le64 rdma_req_tx_memory_mgmt_err;
+	__le64 rsvd52;
+	__le64 rsvd53;
+	__le64 rsvd54;
+
+	/* RDMA/ROCE RESP Error/Debugs (896 - 1023) */
+	__le64 rdma_resp_rx_dup_requests;
+	__le64 rdma_resp_rx_out_of_buffer;
+	__le64 rdma_resp_rx_out_of_seq_pkts;
+	__le64 rdma_resp_rx_cqe_err;
+	__le64 rdma_resp_rx_cqe_flush_err;
+	__le64 rdma_resp_rx_local_len_err;
+	__le64 rdma_resp_rx_inv_request_err;
+	__le64 rdma_resp_rx_local_qp_oper_err;
+
+	__le64 rdma_resp_rx_out_of_atomic_resource;
+	__le64 rdma_resp_tx_pkt_seq_err;
+	__le64 rdma_resp_tx_remote_inv_req_err;
+	__le64 rdma_resp_tx_remote_access_err;
+	__le64 rdma_resp_tx_remote_oper_err;
+	__le64 rdma_resp_tx_rnr_retry_err;
+	__le64 rsvd57;
+	__le64 rsvd58;
+};
+
+/**
+ * struct lif_info - lif info structure
+ */
+struct lif_info {
+	union lif_config config;
+	struct lif_status status;
+	struct lif_stats stats;
+};
+
+union dev_cmd {
+	u32 words[16];
+	struct admin_cmd cmd;
+	struct nop_cmd nop;
+
+	struct dev_identify_cmd identify;
+	struct dev_init_cmd init;
+	struct dev_reset_cmd reset;
+	struct dev_getattr_cmd getattr;
+	struct dev_setattr_cmd setattr;
+
+	struct port_identify_cmd port_identify;
+	struct port_init_cmd port_init;
+	struct port_reset_cmd port_reset;
+	struct port_getattr_cmd port_getattr;
+	struct port_setattr_cmd port_setattr;
+
+	struct lif_identify_cmd lif_identify;
+	struct lif_init_cmd lif_init;
+	struct lif_reset_cmd lif_reset;
+
+	struct qos_identify_cmd qos_identify;
+	struct qos_init_cmd qos_init;
+	struct qos_reset_cmd qos_reset;
+
+	struct q_init_cmd q_init;
+};
+
+union dev_cmd_comp {
+	u32 words[4];
+	u8 status;
+	struct admin_comp comp;
+	struct nop_comp nop;
+
+	struct dev_identify_comp identify;
+	struct dev_init_comp init;
+	struct dev_reset_comp reset;
+	struct dev_getattr_comp getattr;
+	struct dev_setattr_comp setattr;
+
+	struct port_identify_comp port_identify;
+	struct port_init_comp port_init;
+	struct port_reset_comp port_reset;
+	struct port_getattr_comp port_getattr;
+	struct port_setattr_comp port_setattr;
+
+	struct lif_identify_comp lif_identify;
+	struct lif_init_comp lif_init;
+	lif_reset_comp lif_reset;
+
+	struct qos_identify_comp qos_identify;
+	qos_init_comp qos_init;
+	qos_reset_comp qos_reset;
+
+	struct q_init_comp q_init;
+};
+
+/**
+ * union dev_info - Device info register format (read-only)
+ * @signature:       Signature value of 0x44455649 ('DEVI').
+ * @version:         Current version of info.
+ * @asic_type:       Asic type.
+ * @asic_rev:        Asic revision.
+ * @fw_status:       Firmware status.
+ * @fw_heartbeat:    Firmware heartbeat counter.
+ * @serial_num:      Serial number.
+ * @fw_version:      Firmware version.
+ */
+union dev_info_regs {
+#define IONIC_DEVINFO_FWVERS_BUFLEN 32
+#define IONIC_DEVINFO_SERIAL_BUFLEN 32
+	struct {
+		u32    signature;
+		u8     version;
+		u8     asic_type;
+		u8     asic_rev;
+		u8     fw_status;
+		u32    fw_heartbeat;
+		char   fw_version[IONIC_DEVINFO_FWVERS_BUFLEN];
+		char   serial_num[IONIC_DEVINFO_SERIAL_BUFLEN];
+	};
+	u32 words[512];
+};
+
+/**
+ * union dev_cmd_regs - Device command register format (read-write)
+ * @doorbell:        Device Cmd Doorbell, write-only.
+ *                   Write a 1 to signal device to process cmd,
+ *                   poll done for completion.
+ * @done:            Done indicator, bit 0 == 1 when command is complete.
+ * @cmd:             Opcode-specific command bytes
+ * @comp:            Opcode-specific response bytes
+ * @data:            Opcode-specific side-data
+ */
+union dev_cmd_regs {
+	struct {
+		u32                   doorbell;
+		u32                   done;
+		union dev_cmd         cmd;
+		union dev_cmd_comp    comp;
+		u8                    rsvd[48];
+		u32                   data[478];
+	};
+	u32 words[512];
+};
+
+/**
+ * union dev_regs - Device register format in for bar 0 page 0
+ * @info:            Device info registers
+ * @devcmd:          Device command registers
+ */
+union dev_regs {
+	struct {
+		union dev_info_regs info;
+		union dev_cmd_regs  devcmd;
+	};
+	__le32 words[1024];
+};
+
+union adminq_cmd {
+	struct admin_cmd cmd;
+	struct nop_cmd nop;
+	struct q_init_cmd q_init;
+	struct q_control_cmd q_control;
+	struct lif_setattr_cmd lif_setattr;
+	struct lif_getattr_cmd lif_getattr;
+	struct rx_mode_set_cmd rx_mode_set;
+	struct rx_filter_add_cmd rx_filter_add;
+	struct rx_filter_del_cmd rx_filter_del;
+	struct rdma_reset_cmd rdma_reset;
+	struct rdma_queue_cmd rdma_queue;
+	struct fw_download_cmd fw_download;
+	struct fw_control_cmd fw_control;
+};
+
+union adminq_comp {
+	struct admin_comp comp;
+	struct nop_comp nop;
+	struct q_init_comp q_init;
+	struct lif_setattr_comp lif_setattr;
+	struct lif_getattr_comp lif_getattr;
+	struct rx_filter_add_comp rx_filter_add;
+	struct fw_control_comp fw_control;
+};
+
+#define IONIC_BARS_MAX			6
+#define IONIC_PCI_BAR_DBELL		1
+
+/* BAR0 */
+#define BAR0_SIZE			0x8000
+
+#define BAR0_DEV_INFO_REGS_OFFSET	0x0000
+#define BAR0_DEV_CMD_REGS_OFFSET	0x0800
+#define BAR0_DEV_CMD_DATA_REGS_OFFSET	0x0c00
+#define BAR0_INTR_STATUS_OFFSET		0x1000
+#define BAR0_INTR_CTRL_OFFSET		0x2000
+#define DEV_CMD_DONE			0x00000001
+
+#define ASIC_TYPE_CAPRI			0
+
+/**
+ * struct doorbell - Doorbell register layout
+ * @p_index: Producer index
+ * @ring:    Selects the specific ring of the queue to update.
+ *           Type-specific meaning:
+ *              ring=0: Default producer/consumer queue.
+ *              ring=1: (CQ, EQ) Re-Arm queue.  RDMA CQs
+ *              send events to EQs when armed.  EQs send
+ *              interrupts when armed.
+ * @qid:     The queue id selects the queue destination for the
+ *           producer index and flags.
+ */
+struct doorbell {
+	__le16 p_index;
+	u8     ring;
+	u8     qid_lo;
+	__le16 qid_hi;
+	u16    rsvd2;
+};
+
+/**
+ * struct intr_ctrl - Interrupt control register
+ * @coalescing_init:  Coalescing timer initial value, in
+ *                    device units.  Use @identity->intr_coal_mult
+ *                    and @identity->intr_coal_div to convert from
+ *                    usecs to device units:
+ *
+ *                      coal_init = coal_usecs * coal_mutl / coal_div
+ *
+ *                    When an interrupt is sent the interrupt
+ *                    coalescing timer current value
+ *                    (@coalescing_curr) is initialized with this
+ *                    value and begins counting down.  No more
+ *                    interrupts are sent until the coalescing
+ *                    timer reaches 0.  When @coalescing_init=0
+ *                    interrupt coalescing is effectively disabled
+ *                    and every interrupt assert results in an
+ *                    interrupt.  Reset value: 0.
+ * @mask:             Interrupt mask.  When @mask=1 the interrupt
+ *                    resource will not send an interrupt.  When
+ *                    @mask=0 the interrupt resource will send an
+ *                    interrupt if an interrupt event is pending
+ *                    or on the next interrupt assertion event.
+ *                    Reset value: 1.
+ * @int_credits:      Interrupt credits.  This register indicates
+ *                    how many interrupt events the hardware has
+ *                    sent.  When written by software this
+ *                    register atomically decrements @int_credits
+ *                    by the value written.  When @int_credits
+ *                    becomes 0 then the "pending interrupt" bit
+ *                    in the Interrupt Status register is cleared
+ *                    by the hardware and any pending but unsent
+ *                    interrupts are cleared.
+ *                    !!!IMPORTANT!!! This is a signed register.
+ * @flags:            Interrupt control flags
+ *                       @unmask -- When this bit is written with a 1
+ *                       the interrupt resource will set mask=0.
+ *                       @coal_timer_reset -- When this
+ *                       bit is written with a 1 the
+ *                       @coalescing_curr will be reloaded with
+ *                       @coalescing_init to reset the coalescing
+ *                       timer.
+ * @mask_on_assert:   Automatically mask on assertion.  When
+ *                    @mask_on_assert=1 the interrupt resource
+ *                    will set @mask=1 whenever an interrupt is
+ *                    sent.  When using interrupts in Legacy
+ *                    Interrupt mode the driver must select
+ *                    @mask_on_assert=0 for proper interrupt
+ *                    operation.
+ * @coalescing_curr:  Coalescing timer current value, in
+ *                    microseconds.  When this value reaches 0
+ *                    the interrupt resource is again eligible to
+ *                    send an interrupt.  If an interrupt event
+ *                    is already pending when @coalescing_curr
+ *                    reaches 0 the pending interrupt will be
+ *                    sent, otherwise an interrupt will be sent
+ *                    on the next interrupt assertion event.
+ */
+struct intr_ctrl {
+	u8 coalescing_init;
+	u8 rsvd[3];
+	u8 mask;
+	u8 rsvd2[3];
+	u16 int_credits;
+	u16 flags;
+#define INTR_F_UNMASK		0x0001
+#define INTR_F_TIMER_RESET	0x0002
+	u8 mask_on_assert;
+	u8 rsvd3[3];
+	u8 coalescing_curr;
+	u8 rsvd4[3];
+	u32 rsvd6[3];
+};
+
+#define INTR_CTRL_REGS_MAX	2048
+#define INTR_CTRL_COAL_MAX	0x3F
+
+#define intr_to_coal(intr_ctrl)		\
+		((void __iomem *)&(intr_ctrl)->coalescing_init)
+#define intr_to_mask(intr_ctrl)		\
+		((void __iomem *)&(intr_ctrl)->mask)
+#define intr_to_credits(intr_ctrl)	\
+		((void __iomem *)&(intr_ctrl)->int_credits)
+#define intr_to_mask_on_assert(intr_ctrl)\
+		((void __iomem *)&(intr_ctrl)->mask_on_assert)
+
+struct intr_status {
+	u32 status[2];
+};
+
+struct notifyq_cmd {
+	__le32 data;	/* Not used but needed for qcq structure */
+};
+
+union notifyq_comp {
+	struct notifyq_event event;
+	struct link_change_event link_change;
+	struct reset_event reset;
+	struct heartbeat_event heartbeat;
+	struct log_event log;
+};
+
+/* Deprecate */
+struct identity {
+	union drv_identity drv;
+	union dev_identity dev;
+	union lif_identity lif;
+	union port_identity port;
+	union qos_identity qos;
+};
+
+#pragma pack(pop)
+
+#endif /* _IONIC_IF_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_main.c b/drivers/net/ethernet/pensando/ionic/ionic_main.c
index fb49c0ca5095..567ed259b398 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_main.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_main.c
@@ -8,14 +8,309 @@
 
 #include "ionic.h"
 #include "ionic_bus.h"
+#include "ionic_debugfs.h"
 
 MODULE_DESCRIPTION(DRV_DESCRIPTION);
 MODULE_AUTHOR("Pensando Systems, Inc");
 MODULE_LICENSE("GPL");
 MODULE_VERSION(DRV_VERSION);
 
+static const char *ionic_error_to_str(enum status_code code)
+{
+	switch (code) {
+	case IONIC_RC_SUCCESS:
+		return "IONIC_RC_SUCCESS";
+	case IONIC_RC_EVERSION:
+		return "IONIC_RC_EVERSION";
+	case IONIC_RC_EOPCODE:
+		return "IONIC_RC_EOPCODE";
+	case IONIC_RC_EIO:
+		return "IONIC_RC_EIO";
+	case IONIC_RC_EPERM:
+		return "IONIC_RC_EPERM";
+	case IONIC_RC_EQID:
+		return "IONIC_RC_EQID";
+	case IONIC_RC_EQTYPE:
+		return "IONIC_RC_EQTYPE";
+	case IONIC_RC_ENOENT:
+		return "IONIC_RC_ENOENT";
+	case IONIC_RC_EINTR:
+		return "IONIC_RC_EINTR";
+	case IONIC_RC_EAGAIN:
+		return "IONIC_RC_EAGAIN";
+	case IONIC_RC_ENOMEM:
+		return "IONIC_RC_ENOMEM";
+	case IONIC_RC_EFAULT:
+		return "IONIC_RC_EFAULT";
+	case IONIC_RC_EBUSY:
+		return "IONIC_RC_EBUSY";
+	case IONIC_RC_EEXIST:
+		return "IONIC_RC_EEXIST";
+	case IONIC_RC_EINVAL:
+		return "IONIC_RC_EINVAL";
+	case IONIC_RC_ENOSPC:
+		return "IONIC_RC_ENOSPC";
+	case IONIC_RC_ERANGE:
+		return "IONIC_RC_ERANGE";
+	case IONIC_RC_BAD_ADDR:
+		return "IONIC_RC_BAD_ADDR";
+	case IONIC_RC_DEV_CMD:
+		return "IONIC_RC_DEV_CMD";
+	case IONIC_RC_ERROR:
+		return "IONIC_RC_ERROR";
+	case IONIC_RC_ERDMA:
+		return "IONIC_RC_ERDMA";
+	default:
+		return "IONIC_RC_UNKNOWN";
+	}
+}
+
+static int ionic_error_to_errno(enum status_code code)
+{
+	switch (code) {
+	case IONIC_RC_SUCCESS:
+		return 0;
+	case IONIC_RC_EVERSION:
+	case IONIC_RC_EQTYPE:
+	case IONIC_RC_EQID:
+	case IONIC_RC_EINVAL:
+		return -EINVAL;
+	case IONIC_RC_EPERM:
+		return -EPERM;
+	case IONIC_RC_ENOENT:
+		return -ENOENT;
+	case IONIC_RC_EAGAIN:
+		return -EAGAIN;
+	case IONIC_RC_ENOMEM:
+		return -ENOMEM;
+	case IONIC_RC_EFAULT:
+		return -EFAULT;
+	case IONIC_RC_EBUSY:
+		return -EBUSY;
+	case IONIC_RC_EEXIST:
+		return -EEXIST;
+	case IONIC_RC_ENOSPC:
+		return -ENOSPC;
+	case IONIC_RC_ERANGE:
+		return -ERANGE;
+	case IONIC_RC_BAD_ADDR:
+		return -EFAULT;
+	case IONIC_RC_EOPCODE:
+	case IONIC_RC_EINTR:
+	case IONIC_RC_DEV_CMD:
+	case IONIC_RC_ERROR:
+	case IONIC_RC_ERDMA:
+	case IONIC_RC_EIO:
+	default:
+		return -EIO;
+	}
+}
+
+static const char *ionic_opcode_to_str(enum cmd_opcode opcode)
+{
+	switch (opcode) {
+	case CMD_OPCODE_NOP:
+		return "CMD_OPCODE_NOP";
+	case CMD_OPCODE_INIT:
+		return "CMD_OPCODE_INIT";
+	case CMD_OPCODE_RESET:
+		return "CMD_OPCODE_RESET";
+	case CMD_OPCODE_IDENTIFY:
+		return "CMD_OPCODE_IDENTIFY";
+	case CMD_OPCODE_GETATTR:
+		return "CMD_OPCODE_GETATTR";
+	case CMD_OPCODE_SETATTR:
+		return "CMD_OPCODE_SETATTR";
+	case CMD_OPCODE_PORT_IDENTIFY:
+		return "CMD_OPCODE_PORT_IDENTIFY";
+	case CMD_OPCODE_PORT_INIT:
+		return "CMD_OPCODE_PORT_INIT";
+	case CMD_OPCODE_PORT_RESET:
+		return "CMD_OPCODE_PORT_RESET";
+	case CMD_OPCODE_PORT_GETATTR:
+		return "CMD_OPCODE_PORT_GETATTR";
+	case CMD_OPCODE_PORT_SETATTR:
+		return "CMD_OPCODE_PORT_SETATTR";
+	case CMD_OPCODE_LIF_INIT:
+		return "CMD_OPCODE_LIF_INIT";
+	case CMD_OPCODE_LIF_RESET:
+		return "CMD_OPCODE_LIF_RESET";
+	case CMD_OPCODE_LIF_IDENTIFY:
+		return "CMD_OPCODE_LIF_IDENTIFY";
+	case CMD_OPCODE_LIF_SETATTR:
+		return "CMD_OPCODE_LIF_SETATTR";
+	case CMD_OPCODE_LIF_GETATTR:
+		return "CMD_OPCODE_LIF_GETATTR";
+	case CMD_OPCODE_RX_MODE_SET:
+		return "CMD_OPCODE_RX_MODE_SET";
+	case CMD_OPCODE_RX_FILTER_ADD:
+		return "CMD_OPCODE_RX_FILTER_ADD";
+	case CMD_OPCODE_RX_FILTER_DEL:
+		return "CMD_OPCODE_RX_FILTER_DEL";
+	case CMD_OPCODE_Q_INIT:
+		return "CMD_OPCODE_Q_INIT";
+	case CMD_OPCODE_Q_CONTROL:
+		return "CMD_OPCODE_Q_CONTROL";
+	case CMD_OPCODE_RDMA_RESET_LIF:
+		return "CMD_OPCODE_RDMA_RESET_LIF";
+	case CMD_OPCODE_RDMA_CREATE_EQ:
+		return "CMD_OPCODE_RDMA_CREATE_EQ";
+	case CMD_OPCODE_RDMA_CREATE_CQ:
+		return "CMD_OPCODE_RDMA_CREATE_CQ";
+	case CMD_OPCODE_RDMA_CREATE_ADMINQ:
+		return "CMD_OPCODE_RDMA_CREATE_ADMINQ";
+	case CMD_OPCODE_FW_DOWNLOAD:
+		return "CMD_OPCODE_FW_DOWNLOAD";
+	case CMD_OPCODE_FW_CONTROL:
+		return "CMD_OPCODE_FW_CONTROL";
+	default:
+		return "DEVCMD_UNKNOWN";
+	}
+}
+
+int ionic_dev_cmd_wait(struct ionic *ionic, unsigned long max_seconds)
+{
+	struct ionic_dev *idev = &ionic->idev;
+	unsigned long max_wait, start_time, duration;
+	int opcode;
+	int done;
+	int err;
+
+	WARN_ON(in_interrupt());
+
+	/* Wait for dev cmd to complete, retrying if we get EAGAIN,
+	 * but don't wait any longer than max_seconds.
+	 */
+	max_wait = jiffies + (max_seconds * HZ);
+try_again:
+	start_time = jiffies;
+	do {
+		done = ionic_dev_cmd_done(idev);
+		if (done)
+			break;
+		msleep(20);
+	} while (!done && time_before(jiffies, max_wait));
+	duration = jiffies - start_time;
+
+	opcode = idev->dev_cmd_regs->cmd.cmd.opcode;
+	dev_dbg(ionic->dev, "DEVCMD %s (%d) done=%d took %ld secs (%ld jiffies)\n",
+		ionic_opcode_to_str(opcode), opcode,
+		done, duration / HZ, duration);
+
+	if (!done && !time_before(jiffies, max_wait)) {
+		dev_warn(ionic->dev, "DEVCMD %s (%d) timeout after %ld secs\n",
+			 ionic_opcode_to_str(opcode), opcode, max_seconds);
+		return -ETIMEDOUT;
+	}
+
+	err = ionic_dev_cmd_status(&ionic->idev);
+	if (err) {
+		if (err == IONIC_RC_EAGAIN && !time_after(jiffies, max_wait)) {
+			dev_err(ionic->dev, "DEV_CMD %s (%d) error, %s (%d) retrying...\n",
+				ionic_opcode_to_str(opcode), opcode,
+				ionic_error_to_str(err), err);
+
+			msleep(1000);
+			iowrite32(0, &idev->dev_cmd_regs->done);
+			iowrite32(1, &idev->dev_cmd_regs->doorbell);
+			goto try_again;
+		}
+
+		dev_err(ionic->dev, "DEV_CMD %s (%d) error, %s (%d) failed\n",
+			ionic_opcode_to_str(opcode), opcode,
+			ionic_error_to_str(err), err);
+
+		return ionic_error_to_errno(err);
+	}
+
+	return 0;
+}
+
+int ionic_setup(struct ionic *ionic)
+{
+	int err;
+
+	err = ionic_dev_setup(ionic);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+int ionic_identify(struct ionic *ionic)
+{
+	struct identity *ident = &ionic->ident;
+	struct ionic_dev *idev = &ionic->idev;
+	size_t sz;
+	int err;
+
+	memset(ident, 0, sizeof(*ident));
+
+	ident->drv.os_type = cpu_to_le32(IONIC_OS_TYPE_LINUX);
+	ident->drv.os_dist = 0;
+	strncpy(ident->drv.os_dist_str, utsname()->release,
+		sizeof(ident->drv.os_dist_str) - 1);
+	ident->drv.kernel_ver = cpu_to_le32(LINUX_VERSION_CODE);
+	strncpy(ident->drv.kernel_ver_str, utsname()->version,
+		sizeof(ident->drv.kernel_ver_str) - 1);
+	strncpy(ident->drv.driver_ver_str, DRV_VERSION,
+		sizeof(ident->drv.driver_ver_str) - 1);
+
+	mutex_lock(&ionic->dev_cmd_lock);
+
+	sz = min(sizeof(ident->drv), sizeof(idev->dev_cmd_regs->data));
+	memcpy_toio(&idev->dev_cmd_regs->data, &ident->drv, sz);
+
+	ionic_dev_cmd_identify(idev, IONIC_IDENTITY_VERSION_1);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+	if (!err) {
+		sz = min(sizeof(ident->dev), sizeof(idev->dev_cmd_regs->data));
+		memcpy_fromio(&ident->dev, &idev->dev_cmd_regs->data, sz);
+	}
+
+	mutex_unlock(&ionic->dev_cmd_lock);
+
+	if (err)
+		goto err_out_unmap;
+
+	ionic_debugfs_add_ident(ionic);
+
+	return 0;
+
+err_out_unmap:
+	return err;
+}
+
+int ionic_init(struct ionic *ionic)
+{
+	struct ionic_dev *idev = &ionic->idev;
+	int err;
+
+	mutex_lock(&ionic->dev_cmd_lock);
+	ionic_dev_cmd_init(idev);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+	mutex_unlock(&ionic->dev_cmd_lock);
+
+	return err;
+}
+
+int ionic_reset(struct ionic *ionic)
+{
+	struct ionic_dev *idev = &ionic->idev;
+	int err;
+
+	mutex_lock(&ionic->dev_cmd_lock);
+	ionic_dev_cmd_reset(idev);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+	mutex_unlock(&ionic->dev_cmd_lock);
+
+	return err;
+}
+
 static int __init ionic_init_module(void)
 {
+	ionic_struct_size_checks();
+	ionic_debugfs_create();
 	pr_info("%s %s, ver %s\n", DRV_NAME, DRV_DESCRIPTION, DRV_VERSION);
 	return ionic_bus_register_driver();
 }
@@ -23,6 +318,7 @@ static int __init ionic_init_module(void)
 static void __exit ionic_cleanup_module(void)
 {
 	ionic_bus_unregister_driver();
+	ionic_debugfs_destroy();
 }
 
 module_init(ionic_init_module);
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_regs.h b/drivers/net/ethernet/pensando/ionic/ionic_regs.h
new file mode 100644
index 000000000000..3523915061ed
--- /dev/null
+++ b/drivers/net/ethernet/pensando/ionic/ionic_regs.h
@@ -0,0 +1,133 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB OR BSD-2-Clause */
+/* Copyright (c) 2018-2019 Pensando Systems, Inc.  All rights reserved. */
+
+#ifndef IONIC_REGS_H
+#define IONIC_REGS_H
+
+#include <linux/io.h>
+
+/** struct ionic_intr - interrupt control register set.
+ * @coal_init:			coalesce timer initial value.
+ * @mask:			interrupt mask value.
+ * @credits:			interrupt credit count and return.
+ * @mask_assert:		interrupt mask value on assert.
+ * @coal:			coalesce timer time remaining.
+ */
+struct ionic_intr {
+	u32 coal_init;
+	u32 mask;
+	u32 credits;
+	u32 mask_assert;
+	u32 coal;
+	u32 rsvd[3];
+};
+
+/** enum ionic_intr_mask_vals - valid values for mask and mask_assert.
+ * @IONIC_INTR_MASK_CLEAR:	unmask interrupt.
+ * @IONIC_INTR_MASK_SET:	mask interrupt.
+ */
+enum ionic_intr_mask_vals {
+	IONIC_INTR_MASK_CLEAR		= 0,
+	IONIC_INTR_MASK_SET		= 1,
+};
+
+/** enum ionic_intr_credits_bits - bitwise composition of credits values.
+ * @IONIC_INTR_CRED_COUNT:	bit mask of credit count, no shift needed.
+ * @IONIC_INTR_CRED_COUNT_SIGNED: bit mask of credit count, including sign bit.
+ * @IONIC_INTR_CRED_UNMASK:	unmask the interrupt.
+ * @IONIC_INTR_CRED_RESET_COALESCE: reset the coalesce timer.
+ * @IONIC_INTR_CRED_REARM:	unmask the and reset the timer.
+ */
+enum ionic_intr_credits_bits {
+	IONIC_INTR_CRED_COUNT		= 0x7fffu,
+	IONIC_INTR_CRED_COUNT_SIGNED	= 0xffffu,
+	IONIC_INTR_CRED_UNMASK		= 0x10000u,
+	IONIC_INTR_CRED_RESET_COALESCE	= 0x20000u,
+	IONIC_INTR_CRED_REARM		= (IONIC_INTR_CRED_UNMASK |
+					   IONIC_INTR_CRED_RESET_COALESCE),
+};
+
+static inline void ionic_intr_coal_init(struct ionic_intr __iomem *intr_ctrl,
+					int intr_idx, u32 coal)
+{
+	iowrite32(coal, &intr_ctrl[intr_idx].coal_init);
+}
+
+static inline void ionic_intr_mask(struct ionic_intr __iomem *intr_ctrl,
+				   int intr_idx, u32 mask)
+{
+	iowrite32(mask, &intr_ctrl[intr_idx].mask);
+}
+
+static inline void ionic_intr_credits(struct ionic_intr __iomem *intr_ctrl,
+				      int intr_idx, u32 cred, u32 flags)
+{
+	if (WARN_ON_ONCE(cred > IONIC_INTR_CRED_COUNT)) {
+		cred = ioread32(&intr_ctrl[intr_idx].credits);
+		cred &= IONIC_INTR_CRED_COUNT_SIGNED;
+	}
+
+	iowrite32(cred | flags, &intr_ctrl[intr_idx].credits);
+}
+
+static inline void ionic_intr_clean(struct ionic_intr __iomem *intr_ctrl,
+				    int intr_idx)
+{
+	u32 cred;
+
+	cred = ioread32(&intr_ctrl[intr_idx].credits);
+	cred &= IONIC_INTR_CRED_COUNT_SIGNED;
+	cred |= IONIC_INTR_CRED_RESET_COALESCE;
+	iowrite32(cred, &intr_ctrl[intr_idx].credits);
+}
+
+static inline void ionic_intr_mask_assert(struct ionic_intr __iomem *intr_ctrl,
+					  int intr_idx, u32 mask)
+{
+	iowrite32(mask, &intr_ctrl[intr_idx].mask_assert);
+}
+
+/** enum ionic_dbell_bits - bitwise composition of dbell values.
+ *
+ * @IONIC_DBELL_QID_MASK:	unshifted mask of valid queue id bits.
+ * @IONIC_DBELL_QID_SHIFT:	queue id shift amount in dbell value.
+ * @IONIC_DBELL_QID:		macro to build QID component of dbell value.
+ *
+ * @IONIC_DBELL_RING_MASK:	unshifted mask of valid ring bits.
+ * @IONIC_DBELL_RING_SHIFT:	ring shift amount in dbell value.
+ * @IONIC_DBELL_RING:		macro to build ring component of dbell value.
+ *
+ * @IONIC_DBELL_RING_0:		ring zero dbell component value.
+ * @IONIC_DBELL_RING_1:		ring one dbell component value.
+ * @IONIC_DBELL_RING_2:		ring two dbell component value.
+ * @IONIC_DBELL_RING_3:		ring three dbell component value.
+ *
+ * @IONIC_DBELL_INDEX_MASK:	bit mask of valid index bits, no shift needed.
+ */
+enum ionic_dbell_bits {
+	IONIC_DBELL_QID_MASK		= 0xffffff,
+	IONIC_DBELL_QID_SHIFT		= 24,
+
+#define IONIC_DBELL_QID(n) \
+	(((u64)(n) & IONIC_DBELL_QID_MASK) << IONIC_DBELL_QID_SHIFT)
+
+	IONIC_DBELL_RING_MASK		= 0x7,
+	IONIC_DBELL_RING_SHIFT		= 16,
+
+#define IONIC_DBELL_RING(n) \
+	(((u64)(n) & IONIC_DBELL_RING_MASK) << IONIC_DBELL_RING_SHIFT)
+
+	IONIC_DBELL_RING_0		= 0,
+	IONIC_DBELL_RING_1		= IONIC_DBELL_RING(1),
+	IONIC_DBELL_RING_2		= IONIC_DBELL_RING(2),
+	IONIC_DBELL_RING_3		= IONIC_DBELL_RING(3),
+
+	IONIC_DBELL_INDEX_MASK		= 0xffff,
+};
+
+static inline void ionic_dbell_ring(u64 __iomem *db_page, int qtype, u64 val)
+{
+	writeq(val, &db_page[qtype]);
+}
+
+#endif /* IONIC_REGS_H */
-- 
2.17.1


^ permalink raw reply related

* [PATCH v3 net-next 03/19] ionic: Add port management commands
From: Shannon Nelson @ 2019-07-08 19:25 UTC (permalink / raw)
  To: snelson, netdev
In-Reply-To: <20190708192532.27420-1-snelson@pensando.io>

The port management commands apply to the physical port
associated with the PCI device, which might be shared among
several logical interfaces.

Signed-off-by: Shannon Nelson <snelson@pensando.io>
---
 drivers/net/ethernet/pensando/ionic/ionic.h   |   4 +
 .../ethernet/pensando/ionic/ionic_bus_pci.c   |  16 +++
 .../net/ethernet/pensando/ionic/ionic_dev.c   | 116 ++++++++++++++++++
 .../net/ethernet/pensando/ionic/ionic_dev.h   |  15 +++
 .../net/ethernet/pensando/ionic/ionic_main.c  |  95 ++++++++++++++
 5 files changed, 246 insertions(+)

diff --git a/drivers/net/ethernet/pensando/ionic/ionic.h b/drivers/net/ethernet/pensando/ionic/ionic.h
index 4fd0ae86556a..f77712feef5c 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic.h
@@ -42,4 +42,8 @@ int ionic_identify(struct ionic *ionic);
 int ionic_init(struct ionic *ionic);
 int ionic_reset(struct ionic *ionic);
 
+int ionic_port_identify(struct ionic *ionic);
+int ionic_port_init(struct ionic *ionic);
+int ionic_port_reset(struct ionic *ionic);
+
 #endif /* _IONIC_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
index 7eb23f0e87ac..7a41f8995c35 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
@@ -137,8 +137,23 @@ static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto err_out_teardown;
 	}
 
+	/* Configure the ports */
+	err = ionic_port_identify(ionic);
+	if (err) {
+		dev_err(dev, "Cannot identify port: %d, aborting\n", err);
+		goto err_out_reset;
+	}
+
+	err = ionic_port_init(ionic);
+	if (err) {
+		dev_err(dev, "Cannot init port: %d, aborting\n", err);
+		goto err_out_reset;
+	}
+
 	return 0;
 
+err_out_reset:
+	ionic_reset(ionic);
 err_out_teardown:
 	ionic_dev_teardown(ionic);
 err_out_unmap_bars:
@@ -162,6 +177,7 @@ static void ionic_remove(struct pci_dev *pdev)
 	struct ionic *ionic = pci_get_drvdata(pdev);
 
 	if (ionic) {
+		ionic_port_reset(ionic);
 		ionic_reset(ionic);
 		ionic_dev_teardown(ionic);
 		ionic_unmap_bars(ionic);
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.c b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
index ca2ef0a1f620..e2eda2bedac6 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_dev.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.c
@@ -130,3 +130,119 @@ void ionic_dev_cmd_reset(struct ionic_dev *idev)
 
 	ionic_dev_cmd_go(idev, &cmd);
 }
+
+/* Port commands */
+void ionic_dev_cmd_port_identify(struct ionic_dev *idev)
+{
+	union dev_cmd cmd = {
+		.port_init.opcode = CMD_OPCODE_PORT_IDENTIFY,
+		.port_init.index = 0,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_init(struct ionic_dev *idev)
+{
+	union dev_cmd cmd = {
+		.port_init.opcode = CMD_OPCODE_PORT_INIT,
+		.port_init.index = 0,
+		.port_init.info_pa = cpu_to_le64(idev->port_info_pa),
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_reset(struct ionic_dev *idev)
+{
+	union dev_cmd cmd = {
+		.port_reset.opcode = CMD_OPCODE_PORT_RESET,
+		.port_reset.index = 0,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_state(struct ionic_dev *idev, u8 state)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_STATE,
+		.port_setattr.state = state,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_speed(struct ionic_dev *idev, u32 speed)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_SPEED,
+		.port_setattr.speed = cpu_to_le32(speed),
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_mtu(struct ionic_dev *idev, u32 mtu)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_MTU,
+		.port_setattr.mtu = cpu_to_le32(mtu),
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_autoneg(struct ionic_dev *idev, u8 an_enable)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_AUTONEG,
+		.port_setattr.an_enable = an_enable,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_fec(struct ionic_dev *idev, u8 fec_type)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_FEC,
+		.port_setattr.fec_type = fec_type,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_pause(struct ionic_dev *idev, u8 pause_type)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_PAUSE,
+		.port_setattr.pause_type = pause_type,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
+
+void ionic_dev_cmd_port_loopback(struct ionic_dev *idev, u8 loopback_mode)
+{
+	union dev_cmd cmd = {
+		.port_setattr.opcode = CMD_OPCODE_PORT_SETATTR,
+		.port_setattr.index = 0,
+		.port_setattr.attr = IONIC_PORT_ATTR_LOOPBACK,
+		.port_setattr.loopback_mode = loopback_mode,
+	};
+
+	ionic_dev_cmd_go(idev, &cmd);
+}
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.h b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
index 66afab3ee396..fe5e1b0e8d55 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_dev.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.h
@@ -123,6 +123,10 @@ struct ionic_dev {
 	struct ionic_intr __iomem *intr_ctrl;
 	u64 __iomem *intr_status;
 
+	struct port_info *port_info;
+	dma_addr_t port_info_pa;
+	u32 port_info_sz;
+
 	struct ionic_devinfo dev_info;
 };
 
@@ -141,4 +145,15 @@ void ionic_dev_cmd_identify(struct ionic_dev *idev, u8 ver);
 void ionic_dev_cmd_init(struct ionic_dev *idev);
 void ionic_dev_cmd_reset(struct ionic_dev *idev);
 
+void ionic_dev_cmd_port_identify(struct ionic_dev *idev);
+void ionic_dev_cmd_port_init(struct ionic_dev *idev);
+void ionic_dev_cmd_port_reset(struct ionic_dev *idev);
+void ionic_dev_cmd_port_state(struct ionic_dev *idev, u8 state);
+void ionic_dev_cmd_port_speed(struct ionic_dev *idev, u32 speed);
+void ionic_dev_cmd_port_mtu(struct ionic_dev *idev, u32 mtu);
+void ionic_dev_cmd_port_autoneg(struct ionic_dev *idev, u8 an_enable);
+void ionic_dev_cmd_port_fec(struct ionic_dev *idev, u8 fec_type);
+void ionic_dev_cmd_port_pause(struct ionic_dev *idev, u8 pause_type);
+void ionic_dev_cmd_port_loopback(struct ionic_dev *idev, u8 loopback_mode);
+
 #endif /* _IONIC_DEV_H_ */
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_main.c b/drivers/net/ethernet/pensando/ionic/ionic_main.c
index 567ed259b398..78eab326b067 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_main.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_main.c
@@ -307,6 +307,101 @@ int ionic_reset(struct ionic *ionic)
 	return err;
 }
 
+int ionic_port_identify(struct ionic *ionic)
+{
+	struct identity *ident = &ionic->ident;
+	struct ionic_dev *idev = &ionic->idev;
+	struct device *dev = ionic->dev;
+	size_t sz;
+	int err;
+
+	mutex_lock(&ionic->dev_cmd_lock);
+
+	ionic_dev_cmd_port_identify(idev);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+	if (!err) {
+		sz = min(sizeof(ident->port), sizeof(idev->dev_cmd_regs->data));
+		memcpy_fromio(&ident->port, &idev->dev_cmd_regs->data, sz);
+	}
+
+	mutex_unlock(&ionic->dev_cmd_lock);
+
+	dev_dbg(dev, "speed %d\n", ident->port.config.speed);
+	dev_dbg(dev, "mtu %d\n", ident->port.config.mtu);
+	dev_dbg(dev, "state %d\n", ident->port.config.state);
+	dev_dbg(dev, "an_enable %d\n", ident->port.config.an_enable);
+	dev_dbg(dev, "fec_type %d\n", ident->port.config.fec_type);
+	dev_dbg(dev, "pause_type %d\n", ident->port.config.pause_type);
+	dev_dbg(dev, "loopback_mode %d\n", ident->port.config.loopback_mode);
+
+	return err;
+}
+
+int ionic_port_init(struct ionic *ionic)
+{
+	struct identity *ident = &ionic->ident;
+	struct ionic_dev *idev = &ionic->idev;
+	size_t sz;
+	int err;
+
+	if (idev->port_info)
+		return 0;
+
+	idev->port_info_sz = ALIGN(sizeof(*idev->port_info), PAGE_SIZE);
+	idev->port_info = dma_alloc_coherent(ionic->dev, idev->port_info_sz,
+					     &idev->port_info_pa,
+					     GFP_KERNEL);
+	if (!idev->port_info) {
+		dev_err(ionic->dev, "Failed to allocate port info, aborting\n");
+		return -ENOMEM;
+	}
+
+	sz = min(sizeof(ident->port.config), sizeof(idev->dev_cmd_regs->data));
+
+	mutex_lock(&ionic->dev_cmd_lock);
+
+	memcpy_toio(&idev->dev_cmd_regs->data, &ident->port.config, sz);
+	ionic_dev_cmd_port_init(idev);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+
+	ionic_dev_cmd_port_state(&ionic->idev, PORT_ADMIN_STATE_UP);
+	(void)ionic_dev_cmd_wait(ionic, devcmd_timeout);
+
+	mutex_unlock(&ionic->dev_cmd_lock);
+	if (err) {
+		dev_err(ionic->dev, "Failed to init port\n");
+		return err;
+	}
+
+	return 0;
+}
+
+int ionic_port_reset(struct ionic *ionic)
+{
+	struct ionic_dev *idev = &ionic->idev;
+	int err;
+
+	if (!idev->port_info)
+		return 0;
+
+	mutex_lock(&ionic->dev_cmd_lock);
+	ionic_dev_cmd_port_reset(idev);
+	err = ionic_dev_cmd_wait(ionic, devcmd_timeout);
+	mutex_unlock(&ionic->dev_cmd_lock);
+	if (err) {
+		dev_err(ionic->dev, "Failed to reset port\n");
+		return err;
+	}
+
+	dma_free_coherent(ionic->dev, idev->port_info_sz,
+			  idev->port_info, idev->port_info_pa);
+
+	idev->port_info = NULL;
+	idev->port_info_pa = 0;
+
+	return err;
+}
+
 static int __init ionic_init_module(void)
 {
 	ionic_struct_size_checks();
-- 
2.17.1


^ permalink raw reply related

* [PATCH v3 net-next 00/19] Add ionic driver
From: Shannon Nelson @ 2019-07-08 19:25 UTC (permalink / raw)
  To: snelson, netdev

This is a patch series that adds the ionic driver, supporting the Pensando
ethernet device.

In this initial patchset we implement basic transmit and receive.  Later
patchsets will add more advanced features.

Our thanks to Andrew Lunn, Michal Kubecek, Jacub Kicinski, and the ever
present kbuild test robots for their comments and suggestions.

New in v3:
 - use le32_to_cpu() on queue_count[] values in debugfs
 - dma_free_coherent() can handle NULL pointers
 - remove unused SS_TEST from ethtool handlers
 - one more case of stop the tx ring if there is no room
 - remove a couple of stray // comments

New in v2:
 - removed debugfs error checking and cut down on debugfs use
 - remove redundant bounds checking on incoming values for mtu and ethtool
 - don't alloc rx_filter memory until the match type has been checked
 - free the ionic struct on remove
 - simplified link_up and netif_carrier_ok comparison
 - put stats into ethtool -S, out of debugfs
 - moved dev_cmd and dev_info dumping to ethtool -d, out of debugfs
 - added devlink support
 - used kernel's rss init routines rather than open code
 - set the Kbuild dependant on 64BIT
 - cut down on some unnecessary log messaging
 - cleaned up ionic_get_link_ksettings
 - cleaned up other little code bits here and there

Shannon Nelson (19):
  ionic: Add basic framework for IONIC Network device driver
  ionic: Add hardware init and device commands
  ionic: Add port management commands
  ionic: Add basic lif support
  ionic: Add interrupts and doorbells
  ionic: Add basic adminq support
  ionic: Add adminq action
  ionic: Add notifyq support
  ionic: Add the basic NDO callbacks for netdev support
  ionic: Add management of rx filters
  ionic: Add Rx filter and rx_mode ndo support
  ionic: Add async link status check and basic stats
  ionic: Add initial ethtool support
  ionic: Add Tx and Rx handling
  ionic: Add netdev-event handling
  ionic: Add driver stats
  ionic: Add RSS support
  ionic: Add coalesce and other features
  ionic: Add basic devlink interface

 .../networking/device_drivers/index.rst       |    1 +
 .../device_drivers/pensando/ionic.rst         |   64 +
 MAINTAINERS                                   |    8 +
 drivers/net/ethernet/Kconfig                  |    1 +
 drivers/net/ethernet/Makefile                 |    1 +
 drivers/net/ethernet/pensando/Kconfig         |   32 +
 drivers/net/ethernet/pensando/Makefile        |    6 +
 drivers/net/ethernet/pensando/ionic/Makefile  |    8 +
 drivers/net/ethernet/pensando/ionic/ionic.h   |   72 +
 .../net/ethernet/pensando/ionic/ionic_bus.h   |   16 +
 .../ethernet/pensando/ionic/ionic_bus_pci.c   |  291 ++
 .../ethernet/pensando/ionic/ionic_debugfs.c   |  283 ++
 .../ethernet/pensando/ionic/ionic_debugfs.h   |   38 +
 .../net/ethernet/pensando/ionic/ionic_dev.c   |  535 ++++
 .../net/ethernet/pensando/ionic/ionic_dev.h   |  284 ++
 .../ethernet/pensando/ionic/ionic_devlink.c   |   89 +
 .../ethernet/pensando/ionic/ionic_devlink.h   |   12 +
 .../ethernet/pensando/ionic/ionic_ethtool.c   |  788 +++++
 .../ethernet/pensando/ionic/ionic_ethtool.h   |    9 +
 .../net/ethernet/pensando/ionic/ionic_if.h    | 2552 +++++++++++++++++
 .../net/ethernet/pensando/ionic/ionic_lif.c   | 2262 +++++++++++++++
 .../net/ethernet/pensando/ionic/ionic_lif.h   |  270 ++
 .../net/ethernet/pensando/ionic/ionic_main.c  |  553 ++++
 .../net/ethernet/pensando/ionic/ionic_regs.h  |  133 +
 .../ethernet/pensando/ionic/ionic_rx_filter.c |  142 +
 .../ethernet/pensando/ionic/ionic_rx_filter.h |   34 +
 .../net/ethernet/pensando/ionic/ionic_stats.c |  333 +++
 .../net/ethernet/pensando/ionic/ionic_stats.h |   53 +
 .../net/ethernet/pensando/ionic/ionic_txrx.c  |  879 ++++++
 .../net/ethernet/pensando/ionic/ionic_txrx.h  |   15 +
 30 files changed, 9764 insertions(+)
 create mode 100644 Documentation/networking/device_drivers/pensando/ionic.rst
 create mode 100644 drivers/net/ethernet/pensando/Kconfig
 create mode 100644 drivers/net/ethernet/pensando/Makefile
 create mode 100644 drivers/net/ethernet/pensando/ionic/Makefile
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_bus.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_debugfs.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_debugfs.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_dev.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_dev.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_devlink.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_devlink.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_ethtool.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_ethtool.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_if.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_lif.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_lif.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_main.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_regs.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_rx_filter.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_rx_filter.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_stats.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_stats.h
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_txrx.c
 create mode 100644 drivers/net/ethernet/pensando/ionic/ionic_txrx.h

-- 
2.17.1


^ permalink raw reply

* [PATCH v3 2/7] net: phy: realtek: Allow disabling RTL8211E EEE LED mode
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

EEE LED mode is enabled by default on the RTL8211E. Disable it when
the device tree property 'realtek,eee-led-mode-disable' exists.

The magic values to disable EEE LED mode were taken from the RTL8211E
datasheet, unfortunately they are not further documented.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
TODO: DT validation

Changes in v3:
- don't have two versions of rtl8211e_config_init()
  (was due to my dev kernel being 4.19, which doesn't have
   this function yet)
- changed return type of rtl8211e_disable_eee_led_mode() to void
- added empty line after rtl8211e_config_init()

Changes in v2:
- patch added to the series
---
 drivers/net/phy/realtek.c | 31 ++++++++++++++++++++++++++++++-
 1 file changed, 30 insertions(+), 1 deletion(-)

diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index a669945eb829..827ea7ed080d 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -9,8 +9,9 @@
  * Copyright (c) 2004 Freescale Semiconductor, Inc.
  */
 #include <linux/bitops.h>
-#include <linux/phy.h>
 #include <linux/module.h>
+#include <linux/of.h>
+#include <linux/phy.h>
 
 #define RTL821x_PHYSR				0x11
 #define RTL821x_PHYSR_DUPLEX			BIT(13)
@@ -26,6 +27,10 @@
 #define RTL821x_EXT_PAGE_SELECT			0x1e
 #define RTL821x_PAGE_SELECT			0x1f
 
+/* RTL8211E page 5 */
+#define RTL8211E_EEE_LED_MODE1			0x05
+#define RTL8211E_EEE_LED_MODE2			0x06
+
 #define RTL8211F_INSR				0x1d
 
 #define RTL8211F_TX_DELAY			BIT(8)
@@ -53,6 +58,26 @@ static int rtl821x_write_page(struct phy_device *phydev, int page)
 	return __phy_write(phydev, RTL821x_PAGE_SELECT, page);
 }
 
+static void rtl8211e_disable_eee_led_mode(struct phy_device *phydev)
+{
+	int oldpage;
+	int err = 0;
+
+	oldpage = phy_select_page(phydev, 5);
+	if (oldpage < 0)
+		goto out;
+
+	/* write magic values to disable EEE LED mode */
+	err = __phy_write(phydev, RTL8211E_EEE_LED_MODE1, 0x8b82);
+	if (err)
+		goto out;
+
+	err = __phy_write(phydev, RTL8211E_EEE_LED_MODE2, 0x052b);
+
+out:
+	phy_restore_page(phydev, oldpage, err);
+}
+
 static int rtl8201_ack_interrupt(struct phy_device *phydev)
 {
 	int err;
@@ -184,9 +209,13 @@ static int rtl8211f_config_init(struct phy_device *phydev)
 
 static int rtl8211e_config_init(struct phy_device *phydev)
 {
+	struct device *dev = &phydev->mdio.dev;
 	int ret = 0, oldpage;
 	u16 val;
 
+	if (of_property_read_bool(dev->of_node, "realtek,eee-led-mode-disable"))
+		rtl8211e_disable_eee_led_mode(phydev);
+
 	/* enable TX/RX delay for rgmii-* modes, and disable them for rgmii. */
 	switch (phydev->interface) {
 	case PHY_INTERFACE_MODE_RGMII:
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 4/7] net: phy: realtek: Add helpers for accessing RTL8211E extension pages
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

The RTL8211E has extension pages, which can be accessed after
selecting a page through a custom method. Add a function to
modify bits in a register of an extension page and a helper for
selecting an ext page. Use rtl8211e_modify_ext_paged() in
rtl8211e_config_init() instead of doing things 'manually'.

rtl8211e_modify_ext_paged() is inspired by its counterpart
phy_modify_paged().

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
Changes in v3:
- use the new function in rtl8211e_config_init() instead of
  doing things 'manually'
- use existing RTL8211E_EXT_PAGE instead of adding a new define
- updated commit message

Changes in v2:
- use phy_select_page() and phy_restore_page(), get rid of
  rtl8211e_restore_page()
- s/rtl821e_select_ext_page/rtl8211e_select_ext_page/
- updated commit message
---
 drivers/net/phy/realtek.c | 57 +++++++++++++++++++++++++++++----------
 1 file changed, 43 insertions(+), 14 deletions(-)

diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index 827ea7ed080d..fa11ae5ebd91 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -27,6 +27,8 @@
 #define RTL821x_EXT_PAGE_SELECT			0x1e
 #define RTL821x_PAGE_SELECT			0x1f
 
+#define RTL8211E_EXT_PAGE			7
+
 /* RTL8211E page 5 */
 #define RTL8211E_EEE_LED_MODE1			0x05
 #define RTL8211E_EEE_LED_MODE2			0x06
@@ -58,6 +60,44 @@ static int rtl821x_write_page(struct phy_device *phydev, int page)
 	return __phy_write(phydev, RTL821x_PAGE_SELECT, page);
 }
 
+static int rtl8211e_select_ext_page(struct phy_device *phydev, int page)
+{
+	int ret, oldpage;
+
+	oldpage = phy_select_page(phydev, RTL8211E_EXT_PAGE);
+	if (oldpage < 0)
+		return oldpage;
+
+	ret = __phy_write(phydev, RTL821x_EXT_PAGE_SELECT, page);
+	if (ret)
+		return phy_restore_page(phydev, page, ret);
+
+	return 0;
+}
+
+static int rtl8211e_modify_ext_paged(struct phy_device *phydev, int page,
+				     u32 regnum, u16 mask, u16 set)
+{
+	int ret = 0;
+	int oldpage;
+	int new;
+
+	oldpage = rtl8211e_select_ext_page(phydev, page);
+	if (oldpage < 0)
+		goto out;
+
+	ret = __phy_read(phydev, regnum);
+	if (ret < 0)
+		goto out;
+
+	new = (ret & ~mask) | set;
+	if (new != ret)
+		ret = __phy_write(phydev, regnum, new);
+
+out:
+	return phy_restore_page(phydev, oldpage, ret);
+}
+
 static void rtl8211e_disable_eee_led_mode(struct phy_device *phydev)
 {
 	int oldpage;
@@ -210,7 +250,6 @@ static int rtl8211f_config_init(struct phy_device *phydev)
 static int rtl8211e_config_init(struct phy_device *phydev)
 {
 	struct device *dev = &phydev->mdio.dev;
-	int ret = 0, oldpage;
 	u16 val;
 
 	if (of_property_read_bool(dev->of_node, "realtek,eee-led-mode-disable"))
@@ -242,19 +281,9 @@ static int rtl8211e_config_init(struct phy_device *phydev)
 	 * 2 = RX Delay, 1 = TX Delay, 0 = SELRGV (see original PHY datasheet
 	 * for details).
 	 */
-	oldpage = phy_select_page(phydev, 0x7);
-	if (oldpage < 0)
-		goto err_restore_page;
-
-	ret = __phy_write(phydev, RTL821x_EXT_PAGE_SELECT, 0xa4);
-	if (ret)
-		goto err_restore_page;
-
-	ret = __phy_modify(phydev, 0x1c, RTL8211E_TX_DELAY | RTL8211E_RX_DELAY,
-			   val);
-
-err_restore_page:
-	return phy_restore_page(phydev, oldpage, ret);
+	return rtl8211e_modify_ext_paged(phydev, 0xa4, 0x1c,
+					 RTL8211E_TX_DELAY | RTL8211E_RX_DELAY,
+					 val);
 }
 
 static int rtl8211b_suspend(struct phy_device *phydev)
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 7/7] net: phy: realtek: configure RTL8211E LEDs
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

Configure the RTL8211E LEDs behavior when the device tree property
'realtek,led-modes' is specified.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
TODO: DT validation

Changes in v3:
- sanity check led-modes values
- set LACR bits in a more readable way
- use phydev_err() instead of dev_err()
- log an error if LED configuration fails

Changes in v2:
- patch added to the series
---
 drivers/net/phy/realtek.c | 72 +++++++++++++++++++++++++++++++++++++--
 1 file changed, 70 insertions(+), 2 deletions(-)

diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index 5854412403b5..e9fb67654c4e 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -9,10 +9,12 @@
  * Copyright (c) 2004 Freescale Semiconductor, Inc.
  */
 #include <linux/bitops.h>
+#include <linux/bits.h>
 #include <linux/device.h>
 #include <linux/module.h>
 #include <linux/of.h>
 #include <linux/phy.h>
+#include <dt-bindings/net/realtek.h>
 
 #define RTL821x_PHYSR				0x11
 #define RTL821x_PHYSR_DUPLEX			BIT(13)
@@ -34,6 +36,15 @@
 #define RTL8211E_EEE_LED_MODE1			0x05
 #define RTL8211E_EEE_LED_MODE2			0x06
 
+/* RTL8211E extension page 44 */
+#define RTL8211E_LACR				0x1a
+#define RLT8211E_LACR_LEDACTCTRL_SHIFT		4
+#define RLT8211E_LACR_LEDACTCTRL_MASK		GENMASK(6, 4)
+#define RTL8211E_LCR				0x1c
+#define RTL8211E_LCR_LEDCTRL_MASK		(GENMASK(2, 0) | \
+						 GENMASK(6, 4) | \
+						 GENMASK(10, 8))
+
 /* RTL8211E extension page 160 */
 #define RTL8211E_SCR				0x1a
 #define RTL8211E_SCR_DISABLE_RXC_SSC		BIT(2)
@@ -123,6 +134,62 @@ static void rtl8211e_disable_eee_led_mode(struct phy_device *phydev)
 	phy_restore_page(phydev, oldpage, err);
 }
 
+static int rtl8211e_config_leds(struct phy_device *phydev)
+{
+	struct device *dev = &phydev->mdio.dev;
+	int count, i, oldpage, ret;
+	u16 lacr_bits = 0, lcr_bits = 0;
+
+	if (!dev->of_node)
+		return 0;
+
+	if (of_property_read_bool(dev->of_node, "realtek,eee-led-mode-disable"))
+		rtl8211e_disable_eee_led_mode(phydev);
+
+	count = of_property_count_elems_of_size(dev->of_node,
+						"realtek,led-modes",
+						sizeof(u32));
+	if (count < 0 || count > 3)
+		return -EINVAL;
+
+	for (i = 0; i < count; i++) {
+		u32 val;
+
+		of_property_read_u32_index(dev->of_node,
+					   "realtek,led-modes", i, &val);
+		if ((val > RTL8211E_LINK_10_100_1000 &&
+		    val < RTL8211E_LINK_ACTIVITY) ||
+		    val > (RTL8211E_LINK_ACTIVITY | RTL8211E_LINK_10_100_1000))
+			return -EINVAL;
+
+		if (val & RTL8211E_LINK_ACTIVITY)
+			lacr_bits |= BIT(RLT8211E_LACR_LEDACTCTRL_SHIFT + i);
+
+		lcr_bits |= (u16)(val & 0xf) << (i * 4);
+	}
+
+	oldpage = rtl8211e_select_ext_page(phydev, 44);
+	if (oldpage < 0) {
+		phydev_err(phydev, "failed to select extended page: %d\n", oldpage);
+		goto err;
+	}
+
+	ret = __phy_modify(phydev, RTL8211E_LACR,
+			   RLT8211E_LACR_LEDACTCTRL_MASK, lacr_bits);
+	if (ret) {
+		phydev_err(phydev, "failed to write LACR reg: %d\n", ret);
+		goto err;
+	}
+
+	ret = __phy_modify(phydev, RTL8211E_LCR,
+			   RTL8211E_LCR_LEDCTRL_MASK, lcr_bits);
+	if (ret)
+		phydev_err(phydev, "failed to write LCR reg: %d\n", ret);
+
+err:
+	return phy_restore_page(phydev, oldpage, ret);
+}
+
 static int rtl8201_ack_interrupt(struct phy_device *phydev)
 {
 	int err;
@@ -267,8 +334,9 @@ static int rtl8211e_config_init(struct phy_device *phydev)
 				   ret);
 	}
 
-	if (of_property_read_bool(dev->of_node, "realtek,eee-led-mode-disable"))
-		rtl8211e_disable_eee_led_mode(phydev);
+	ret = rtl8211e_config_leds(phydev);
+	if (ret)
+		phydev_err(phydev, "LED configuration failed: %d\n", ret);
 
 	/* enable TX/RX delay for rgmii-* modes, and disable them for rgmii. */
 	switch (phydev->interface) {
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 6/7] dt-bindings: net: realtek: Add property to configure LED mode
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

The LED behavior of some Realtek PHYs is configurable. Add the
property 'realtek,led-modes' to specify the configuration of the
LEDs.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
Changes in v3:
- added RTL8211E_LED_OFF to LED modes
- changed wording for supported PHY models

Changes in v2:
- patch added to the series
---
 .../devicetree/bindings/net/realtek.txt        | 11 +++++++++++
 include/dt-bindings/net/realtek.h              | 18 ++++++++++++++++++
 2 files changed, 29 insertions(+)
 create mode 100644 include/dt-bindings/net/realtek.h

diff --git a/Documentation/devicetree/bindings/net/realtek.txt b/Documentation/devicetree/bindings/net/realtek.txt
index af2824664f08..ab218f2b7653 100644
--- a/Documentation/devicetree/bindings/net/realtek.txt
+++ b/Documentation/devicetree/bindings/net/realtek.txt
@@ -19,6 +19,14 @@ Optional properties:
 
   Only supported for "realtek,rtl8211e".
 
+- realtek,led-modes: LED mode configuration.
+
+  A 0..3 element vector, with each element configuring the operating
+  mode of an LED. Omitted LEDs are turned off. Allowed values are
+  defined in "include/dt-bindings/net/realtek.h".
+
+  Only supported for "realtek,rtl8211e".
+
 
 Example:
 
@@ -32,5 +40,8 @@ mdio0 {
 		reg = <1>;
 		realtek,eee-led-mode-disable;
 		realtek,enable-ssc;
+		realtek,led-modes = <RTL8211E_LINK_ACTIVITY
+				     RTL8211E_LINK_100
+				     RTL8211E_LINK_1000>;
 	};
 };
diff --git a/include/dt-bindings/net/realtek.h b/include/dt-bindings/net/realtek.h
new file mode 100644
index 000000000000..023cb776a7a3
--- /dev/null
+++ b/include/dt-bindings/net/realtek.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _DT_BINDINGS_REALTEK_H
+#define _DT_BINDINGS_REALTEK_H
+
+/* LED modes for RTL8211E PHY */
+
+#define RTL8211E_LED_OFF		0
+#define RTL8211E_LINK_10		1
+#define RTL8211E_LINK_100		2
+#define RTL8211E_LINK_1000		4
+#define RTL8211E_LINK_10_100		3
+#define RTL8211E_LINK_10_1000		5
+#define RTL8211E_LINK_100_1000		6
+#define RTL8211E_LINK_10_100_1000	7
+
+#define RTL8211E_LINK_ACTIVITY		(1 << 16)
+
+#endif
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 5/7] net: phy: realtek: Support SSC for the RTL8211E
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

By default Spread-Spectrum Clocking (SSC) is disabled on the RTL8211E.
Enable it if the device tree property 'realtek,enable-ssc' exists.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
TODO: DT validation

Changes in v3:
- use phydev_err() instead of dev_err()

Changes in v2:
- enable SSC in config_init() instead of probe()
- fixed error check after enabling SSC
---
 drivers/net/phy/realtek.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index fa11ae5ebd91..5854412403b5 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -9,6 +9,7 @@
  * Copyright (c) 2004 Freescale Semiconductor, Inc.
  */
 #include <linux/bitops.h>
+#include <linux/device.h>
 #include <linux/module.h>
 #include <linux/of.h>
 #include <linux/phy.h>
@@ -33,6 +34,10 @@
 #define RTL8211E_EEE_LED_MODE1			0x05
 #define RTL8211E_EEE_LED_MODE2			0x06
 
+/* RTL8211E extension page 160 */
+#define RTL8211E_SCR				0x1a
+#define RTL8211E_SCR_DISABLE_RXC_SSC		BIT(2)
+
 #define RTL8211F_INSR				0x1d
 
 #define RTL8211F_TX_DELAY			BIT(8)
@@ -250,8 +255,18 @@ static int rtl8211f_config_init(struct phy_device *phydev)
 static int rtl8211e_config_init(struct phy_device *phydev)
 {
 	struct device *dev = &phydev->mdio.dev;
+	int ret;
 	u16 val;
 
+	if (of_property_read_bool(dev->of_node, "realtek,enable-ssc")) {
+		ret = rtl8211e_modify_ext_paged(phydev, 0xa0, RTL8211E_SCR,
+						RTL8211E_SCR_DISABLE_RXC_SSC,
+						0);
+		if (ret < 0)
+			phydev_err(phydev, "failed to enable SSC on RXC: %d\n",
+				   ret);
+	}
+
 	if (of_property_read_bool(dev->of_node, "realtek,eee-led-mode-disable"))
 		rtl8211e_disable_eee_led_mode(phydev);
 
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 3/7] dt-bindings: net: realtek: Add property to enable SSC
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke
In-Reply-To: <20190708192459.187984-1-mka@chromium.org>

Add the 'realtek,enable-ssc' property to enable Spread Spectrum
Clocking (SSC) on Realtek PHYs that support it.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
Changes in v3:
- changed wording for supported PHY models

Changes in v2:
- patch added to the series (kind of, it already existed, but now
  the binding is created by another patch)
---
 Documentation/devicetree/bindings/net/realtek.txt | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/Documentation/devicetree/bindings/net/realtek.txt b/Documentation/devicetree/bindings/net/realtek.txt
index db0333f23fec..af2824664f08 100644
--- a/Documentation/devicetree/bindings/net/realtek.txt
+++ b/Documentation/devicetree/bindings/net/realtek.txt
@@ -15,6 +15,10 @@ Optional properties:
 
   Only supported for "realtek,rtl8211e".
 
+- realtek,enable-ssc : Enable Spread Spectrum Clocking (SSC) on this port.
+
+  Only supported for "realtek,rtl8211e".
+
 
 Example:
 
@@ -27,5 +31,6 @@ mdio0 {
 		compatible = "realtek,rtl8211e";
 		reg = <1>;
 		realtek,eee-led-mode-disable;
+		realtek,enable-ssc;
 	};
 };
-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply related

* [PATCH v3 0/7] net: phy: realtek: Enable configuration of RTL8211E LEDs and SSC
From: Matthias Kaehlcke @ 2019-07-08 19:24 UTC (permalink / raw)
  To: David S . Miller, Rob Herring, Mark Rutland, Andrew Lunn,
	Florian Fainelli, Heiner Kallweit
  Cc: netdev, devicetree, linux-kernel, Douglas Anderson,
	Matthias Kaehlcke

The Realtek RTL8211E allows customization of the PHY LED behavior,
like which LEDs are on for certain link speeds and which LEDs blink
when there is traffic. By default EEE LED mode is enabled, in which
a blinking LED is on for 400ms and off for 2s. This series adds
support for configuring the LED behavior through device tree
properties.

The RTL8211E supports Spread Spectrum Clocking (SSC), which reduces
clock noise that may affect other board functions. By default SSC
is disabled, this series adds support for enabling it through a
device tree property.

Certain registers on the RTL8211E can only be accessed through
a vendor specific extended page mechanism. Extended pages need
to be accessed for the LED configuration and enabling SSC. This
series adds helpers to facilitate accessing extended pages.

Matthias Kaehlcke (7):
  dt-bindings: net: Add bindings for Realtek PHYs
  net: phy: realtek: Allow disabling RTL8211E EEE LED mode
  dt-bindings: net: realtek: Add property to enable SSC
  net: phy: realtek: Add helpers for accessing RTL8211E extension pages
  net: phy: realtek: Support SSC for the RTL8211E
  dt-bindings: net: realtek: Add property to configure LED mode
  net: phy: realtek: configure RTL8211E LEDs

 .../devicetree/bindings/net/realtek.txt       |  47 +++++
 drivers/net/phy/realtek.c                     | 171 ++++++++++++++++--
 include/dt-bindings/net/realtek.h             |  18 ++
 3 files changed, 221 insertions(+), 15 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/net/realtek.txt
 create mode 100644 include/dt-bindings/net/realtek.h

-- 
2.22.0.410.gd8fdbe21b5-goog


^ permalink raw reply

* Re: [PATCH net-next] sfc: Remove 'PCIE error reporting unavailable'
From: David Miller @ 2019-07-08 19:17 UTC (permalink / raw)
  To: mhabets; +Cc: linux-net-drivers, netdev
In-Reply-To: <156258403191.17195.13184667600147687856.stgit@mh-desktop.uk.solarflarecom.com>

From: Martin Habets <mhabets@solarflare.com>
Date: Mon, 8 Jul 2019 12:07:11 +0100

> This is only at notice level but it was pointed out that no other driver
> does this.
> Also there is no action the user can take as it is really a property of
> the server.
> 
> Signed-off-by: Martin Habets <mhabets@solarflare.com>

Applied, thanks.

^ permalink raw reply

* [bpf PATCH v2 6/6] bpf: sockmap/tls, close can race with map free
From: John Fastabend @ 2019-07-08 19:15 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

When a map free is called and in parallel a socket is closed we
have two paths that can potentially reset the socket prot ops, the
bpf close() path and the map free path. This creates a problem
with which prot ops should be used from the socket closed side.

If the map_free side completes first then we want to call the
original lowest level ops. However, if the tls path runs first
we want to call the sockmap ops. Additionally there was no locking
around prot updates in TLS code paths so the prot ops could
be changed multiple times once from TLS path and again from sockmap
side potentially leaving ops pointed at either TLS or sockmap
when psock and/or tls context have already been destroyed.

To fix this race first only update ops inside callback lock
so that TLS, sockmap and lowest level all agree on prot state.
Second and a ULP callback update() so that lower layers can
inform the upper layer when they are being removed allowing the
upper layer to reset prot ops.

This gets us close to allowing sockmap and tls to be stacked
in arbitrary order but will save that patch for *next trees.

Reported-by: syzbot+06537213db7ba2745c4a@syzkaller.appspotmail.com
Fixes: 02c558b2d5d6 ("bpf: sockmap, support for msg_peek in sk_msg with redirect ingress")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 include/linux/skmsg.h |    8 +++++++-
 include/net/tcp.h     |    3 +++
 net/core/skmsg.c      |    4 ++--
 net/ipv4/tcp_ulp.c    |   13 +++++++++++++
 net/tls/tls_main.c    |   35 +++++++++++++++++++++++++++++------
 5 files changed, 54 insertions(+), 9 deletions(-)

diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h
index 50ced8aba9db..e4b3fb4bb77c 100644
--- a/include/linux/skmsg.h
+++ b/include/linux/skmsg.h
@@ -354,7 +354,13 @@ static inline void sk_psock_restore_proto(struct sock *sk,
 	sk->sk_write_space = psock->saved_write_space;
 
 	if (psock->sk_proto) {
-		sk->sk_prot = psock->sk_proto;
+		struct inet_connection_sock *icsk = inet_csk(sk);
+		bool has_ulp = !!icsk->icsk_ulp_data;
+
+		if (has_ulp)
+			tcp_update_ulp(sk, psock->sk_proto);
+		else
+			sk->sk_prot = psock->sk_proto;
 		psock->sk_proto = NULL;
 	}
 }
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 9d36cc88d043..123cac4c96f2 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -2102,6 +2102,8 @@ struct tcp_ulp_ops {
 
 	/* initialize ulp */
 	int (*init)(struct sock *sk);
+	/* update ulp */
+	void (*update)(struct sock *sk, struct proto *p);
 	/* cleanup ulp */
 	void (*release)(struct sock *sk);
 
@@ -2113,6 +2115,7 @@ void tcp_unregister_ulp(struct tcp_ulp_ops *type);
 int tcp_set_ulp(struct sock *sk, const char *name);
 void tcp_get_available_ulp(char *buf, size_t len);
 void tcp_cleanup_ulp(struct sock *sk);
+void tcp_update_ulp(struct sock *sk, struct proto *p);
 
 #define MODULE_ALIAS_TCP_ULP(name)				\
 	__MODULE_INFO(alias, alias_userspace, name);		\
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index 93bffaad2135..6832eeb4b785 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -585,12 +585,12 @@ EXPORT_SYMBOL_GPL(sk_psock_destroy);
 
 void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
 {
-	rcu_assign_sk_user_data(sk, NULL);
 	sk_psock_cork_free(psock);
 	sk_psock_zap_ingress(psock);
-	sk_psock_restore_proto(sk, psock);
 
 	write_lock_bh(&sk->sk_callback_lock);
+	sk_psock_restore_proto(sk, psock);
+	rcu_assign_sk_user_data(sk, NULL);
 	if (psock->progs.skb_parser)
 		sk_psock_stop_strp(sk, psock);
 	write_unlock_bh(&sk->sk_callback_lock);
diff --git a/net/ipv4/tcp_ulp.c b/net/ipv4/tcp_ulp.c
index 3d8a1d835471..4849edb62d52 100644
--- a/net/ipv4/tcp_ulp.c
+++ b/net/ipv4/tcp_ulp.c
@@ -96,6 +96,19 @@ void tcp_get_available_ulp(char *buf, size_t maxlen)
 	rcu_read_unlock();
 }
 
+void tcp_update_ulp(struct sock *sk, struct proto *proto)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	if (!icsk->icsk_ulp_ops) {
+		sk->sk_prot = proto;
+		return;
+	}
+
+	if (icsk->icsk_ulp_ops->update)
+		icsk->icsk_ulp_ops->update(sk, proto);
+}
+
 void tcp_cleanup_ulp(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index e8418456ee24..4ba5476fbc5f 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -336,15 +336,17 @@ static void tls_sk_proto_unhash(struct sock *sk)
 
 static void tls_sk_proto_close(struct sock *sk, long timeout)
 {
+	struct inet_connection_sock *icsk = inet_csk(sk);
 	struct tls_context *ctx = tls_get_ctx(sk);
 	long timeo = sock_sndtimeo(sk, 0);
-	void (*sk_proto_close)(struct sock *sk, long timeout);
+
+	if (unlikely(!ctx))
+		return;
 
 	if (ctx->tx_conf == TLS_SW)
 		tls_sw_cancel_work_tx(ctx);
 
 	lock_sock(sk);
-	sk_proto_close = ctx->sk_proto_close;
 
 	if (ctx->tx_conf == TLS_HW_RECORD && ctx->rx_conf == TLS_HW_RECORD)
 		goto skip_tx_cleanup;
@@ -352,15 +354,18 @@ static void tls_sk_proto_close(struct sock *sk, long timeout)
 	if (ctx->tx_conf == TLS_BASE && ctx->rx_conf == TLS_BASE)
 		goto skip_tx_cleanup;
 
-	sk->sk_prot = ctx->sk_proto;
 	tls_sk_proto_cleanup(sk, ctx, timeo);
 
 skip_tx_cleanup:
+	write_lock_bh(&sk->sk_callback_lock);
+	icsk->icsk_ulp_data = NULL;
+	if (sk->sk_prot->close == tls_sk_proto_close)
+		sk->sk_prot = ctx->sk_proto;
+	write_unlock_bh(&sk->sk_callback_lock);
 	release_sock(sk);
 	if (ctx->rx_conf == TLS_SW)
 		tls_sw_release_strp_rx(ctx);
-	sk_proto_close(sk, timeout);
-
+	ctx->sk_proto_close(sk, timeout);
 	if (ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW &&
 	    ctx->tx_conf != TLS_HW_RECORD && ctx->rx_conf != TLS_HW_RECORD)
 		tls_ctx_free(ctx);
@@ -836,22 +841,39 @@ static int tls_init(struct sock *sk)
 	if (sk->sk_state != TCP_ESTABLISHED)
 		return -ENOTSUPP;
 
+	tls_build_proto(sk);
+
 	/* allocate tls context */
+	write_lock_bh(&sk->sk_callback_lock);
 	ctx = create_ctx(sk);
 	if (!ctx) {
 		rc = -ENOMEM;
 		goto out;
 	}
 
-	tls_build_proto(sk);
 	ctx->tx_conf = TLS_BASE;
 	ctx->rx_conf = TLS_BASE;
 	ctx->sk_proto = sk->sk_prot;
 	update_sk_prot(sk, ctx);
 out:
+	write_unlock_bh(&sk->sk_callback_lock);
 	return rc;
 }
 
+static void tls_update(struct sock *sk, struct proto *p)
+{
+	struct tls_context *ctx;
+
+	ctx = tls_get_ctx(sk);
+	if (likely(ctx)) {
+		ctx->sk_proto_close = p->close;
+		ctx->unhash = p->unhash;
+		ctx->sk_proto = p;
+	} else {
+		sk->sk_prot = p;
+	}
+}
+
 void tls_register_device(struct tls_device *device)
 {
 	spin_lock_bh(&device_spinlock);
@@ -872,6 +894,7 @@ static struct tcp_ulp_ops tcp_tls_ulp_ops __read_mostly = {
 	.name			= "tls",
 	.owner			= THIS_MODULE,
 	.init			= tls_init,
+	.update			= tls_update,
 };
 
 static int __init tls_register(void)


^ permalink raw reply related

* [bpf PATCH v2 5/6] bpf: sockmap, only create entry if ulp is not already enabled
From: John Fastabend @ 2019-07-08 19:15 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

Sockmap does not currently support adding sockets after TLS has been
enabled. There never was a real use case for this so it was never
added. But, we lost the test for ULP at some point so add it here
and fail the socket insert if TLS is enabled. Future work could
make sockmap support this use case but fixup the bug here.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 net/core/sock_map.c |    3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 56bcabe7c2f2..1330a7442e5b 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -334,6 +334,7 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
 				  struct sock *sk, u64 flags)
 {
 	struct bpf_stab *stab = container_of(map, struct bpf_stab, map);
+	struct inet_connection_sock *icsk = inet_csk(sk);
 	struct sk_psock_link *link;
 	struct sk_psock *psock;
 	struct sock *osk;
@@ -344,6 +345,8 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
 		return -EINVAL;
 	if (unlikely(idx >= map->max_entries))
 		return -E2BIG;
+	if (unlikely(icsk->icsk_ulp_data))
+		return -EINVAL;
 
 	link = sk_psock_init_link();
 	if (!link)


^ permalink raw reply related

* [bpf PATCH v2 4/6] bpf: sockmap, synchronize_rcu before free'ing map
From: John Fastabend @ 2019-07-08 19:14 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

We need to have a synchronize_rcu before free'ing the sockmap because
any outstanding psock references will have a pointer to the map and
when they use this could trigger a use after free.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 net/core/sock_map.c |    2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 28702f2e9a4a..56bcabe7c2f2 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -247,6 +247,8 @@ static void sock_map_free(struct bpf_map *map)
 	raw_spin_unlock_bh(&stab->lock);
 	rcu_read_unlock();
 
+	synchronize_rcu();
+
 	bpf_map_area_free(stab->sks);
 	kfree(stab);
 }


^ permalink raw reply related

* [bpf PATCH v2 3/6] bpf: sockmap, sock_map_delete needs to use xchg
From: John Fastabend @ 2019-07-08 19:14 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

__sock_map_delete() may be called from a tcp event such as unhash or
close from the following trace,

  tcp_bpf_close()
    tcp_bpf_remove()
      sk_psock_unlink()
        sock_map_delete_from_link()
          __sock_map_delete()

In this case the sock lock is held but this only protects against
duplicate removals on the TCP side. If the map is free'd then we have
this trace,

  sock_map_free
    xchg()                  <- replaces map entry
    sock_map_unref()
      sk_psock_put()
        sock_map_del_link()

The __sock_map_delete() call however uses a read, test, null over the
map entry which can result in both paths trying to free the map
entry.

To fix use xchg in TCP paths as well so we avoid having two references
to the same map entry.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 net/core/sock_map.c |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 52d4faeee18b..28702f2e9a4a 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -276,16 +276,20 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
 			     struct sock **psk)
 {
 	struct sock *sk;
+	int err = 0;
 
 	raw_spin_lock_bh(&stab->lock);
 	sk = *psk;
 	if (!sk_test || sk_test == sk)
-		*psk = NULL;
+		sk = xchg(psk, NULL);
+
+	if (likely(sk))
+		sock_map_unref(sk, psk);
+	else
+		err = -EINVAL;
+
 	raw_spin_unlock_bh(&stab->lock);
-	if (unlikely(!sk))
-		return -EINVAL;
-	sock_map_unref(sk, psk);
-	return 0;
+	return err;
 }
 
 static void sock_map_delete_from_link(struct bpf_map *map, struct sock *sk,


^ permalink raw reply related

* [bpf PATCH v2 2/6] bpf: tls fix transition through disconnect with close
From: John Fastabend @ 2019-07-08 19:14 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

It is possible (via shutdown()) for TCP socks to go through TCP_CLOSE
state via tcp_dosconnect() without actually calling tcp_close which
would then call the tls close callback. Because of this a user could
disconnect a socket then put it in a LISTEN state which would break
our assumptions about sockets always being ESTABLISHED state.

More directly because close() can call unhash() and unhash is
implemented by sockmap if a sockmap socket has TLS enabled we can
incorrectly destroy the psock from unhash() and then call its close
handler again. But because the psock (sockmap socket representation)
is already destroyed we call close handler in sk->prot. However,
in some cases (TLS BASE/BASE case) this will still point at the
sockmap close handler resulting in a circular call and crash reported
by syzbot.

To fix both above issues implement the unhash() routine for TLS.

Fixes: 3c4d7559159bf ("tls: kernel TLS support")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 include/net/tls.h  |    6 +++++-
 net/tls/tls_main.c |   54 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 59 insertions(+), 1 deletion(-)

diff --git a/include/net/tls.h b/include/net/tls.h
index 0a3540a6304d..2a98d0584999 100644
--- a/include/net/tls.h
+++ b/include/net/tls.h
@@ -251,6 +251,8 @@ struct tls_context {
 	u8 tx_conf:3;
 	u8 rx_conf:3;
 
+	struct proto *sk_proto;
+
 	int (*push_pending_record)(struct sock *sk, int flags);
 	void (*sk_write_space)(struct sock *sk);
 
@@ -288,6 +290,8 @@ struct tls_context {
 
 	struct list_head list;
 	refcount_t refcount;
+
+	struct work_struct gc;
 };
 
 enum tls_offload_ctx_dir {
@@ -356,7 +360,7 @@ int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx, int tx);
 int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size);
 int tls_sw_sendpage(struct sock *sk, struct page *page,
 		    int offset, size_t size, int flags);
-void tls_sw_close(struct sock *sk, long timeout);
+void tls_sw_cancel_work_tx(struct tls_context *tls_ctx);
 void tls_sw_free_resources_tx(struct sock *sk);
 void tls_sw_free_resources_rx(struct sock *sk);
 void tls_sw_release_resources_rx(struct sock *sk);
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index d63c3583d2f7..e8418456ee24 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -251,6 +251,32 @@ static void tls_write_space(struct sock *sk)
 	ctx->sk_write_space(sk);
 }
 
+static void tls_ctx_free_deferred(struct work_struct *gc)
+{
+	struct tls_context *ctx = container_of(gc, struct tls_context, gc);
+
+	if (ctx->rx_conf == TLS_SW)
+		tls_sw_release_strp_rx(ctx);
+
+	/* Ensure any remaining work items are completed. The sk will
+	 * already have lost its tls_ctx reference by the time we get
+	 * here so no xmit operation will actually be performed.
+	 */
+	tls_sw_cancel_work_tx(ctx);
+	kfree(ctx);
+}
+
+static void tls_ctx_free_wq(struct tls_context *ctx)
+{
+	if (!ctx)
+		return;
+
+	memzero_explicit(&ctx->crypto_send, sizeof(ctx->crypto_send));
+	memzero_explicit(&ctx->crypto_recv, sizeof(ctx->crypto_recv));
+	INIT_WORK(&ctx->gc, tls_ctx_free_deferred);
+	schedule_work(&ctx->gc);
+}
+
 static void tls_ctx_free(struct tls_context *ctx)
 {
 	if (!ctx)
@@ -287,6 +313,27 @@ static void tls_sk_proto_cleanup(struct sock *sk,
 #endif
 }
 
+static void tls_sk_proto_unhash(struct sock *sk)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	long timeo = sock_sndtimeo(sk, 0);
+	struct tls_context *ctx;
+
+	if (unlikely(!icsk->icsk_ulp_data)) {
+		if (sk->sk_prot->unhash)
+			sk->sk_prot->unhash(sk);
+	}
+
+	ctx = tls_get_ctx(sk);
+	if (ctx->tx_conf == TLS_SW || ctx->rx_conf == TLS_SW)
+		tls_sk_proto_cleanup(sk, ctx, timeo);
+	icsk->icsk_ulp_data = NULL;
+	tls_ctx_free_wq(ctx);
+
+	if (ctx->unhash)
+		ctx->unhash(sk);
+}
+
 static void tls_sk_proto_close(struct sock *sk, long timeout)
 {
 	struct tls_context *ctx = tls_get_ctx(sk);
@@ -305,6 +352,7 @@ static void tls_sk_proto_close(struct sock *sk, long timeout)
 	if (ctx->tx_conf == TLS_BASE && ctx->rx_conf == TLS_BASE)
 		goto skip_tx_cleanup;
 
+	sk->sk_prot = ctx->sk_proto;
 	tls_sk_proto_cleanup(sk, ctx, timeo);
 
 skip_tx_cleanup:
@@ -606,6 +654,7 @@ static struct tls_context *create_ctx(struct sock *sk)
 	ctx->setsockopt = sk->sk_prot->setsockopt;
 	ctx->getsockopt = sk->sk_prot->getsockopt;
 	ctx->sk_proto_close = sk->sk_prot->close;
+	ctx->unhash = sk->sk_prot->unhash;
 	return ctx;
 }
 
@@ -729,20 +778,24 @@ static void build_protos(struct proto prot[TLS_NUM_CONFIG][TLS_NUM_CONFIG],
 	prot[TLS_BASE][TLS_BASE].setsockopt	= tls_setsockopt;
 	prot[TLS_BASE][TLS_BASE].getsockopt	= tls_getsockopt;
 	prot[TLS_BASE][TLS_BASE].close		= tls_sk_proto_close;
+	prot[TLS_BASE][TLS_BASE].unhash		= tls_sk_proto_unhash;
 
 	prot[TLS_SW][TLS_BASE] = prot[TLS_BASE][TLS_BASE];
 	prot[TLS_SW][TLS_BASE].sendmsg		= tls_sw_sendmsg;
 	prot[TLS_SW][TLS_BASE].sendpage		= tls_sw_sendpage;
+	prot[TLS_SW][TLS_BASE].unhash		= tls_sk_proto_unhash;
 
 	prot[TLS_BASE][TLS_SW] = prot[TLS_BASE][TLS_BASE];
 	prot[TLS_BASE][TLS_SW].recvmsg		  = tls_sw_recvmsg;
 	prot[TLS_BASE][TLS_SW].stream_memory_read = tls_sw_stream_read;
 	prot[TLS_BASE][TLS_SW].close		  = tls_sk_proto_close;
+	prot[TLS_BASE][TLS_SW].unhash		  = tls_sk_proto_unhash;
 
 	prot[TLS_SW][TLS_SW] = prot[TLS_SW][TLS_BASE];
 	prot[TLS_SW][TLS_SW].recvmsg		= tls_sw_recvmsg;
 	prot[TLS_SW][TLS_SW].stream_memory_read	= tls_sw_stream_read;
 	prot[TLS_SW][TLS_SW].close		= tls_sk_proto_close;
+	prot[TLS_SW][TLS_SW].unhash		= tls_sk_proto_unhash;
 
 #ifdef CONFIG_TLS_DEVICE
 	prot[TLS_HW][TLS_BASE] = prot[TLS_BASE][TLS_BASE];
@@ -793,6 +846,7 @@ static int tls_init(struct sock *sk)
 	tls_build_proto(sk);
 	ctx->tx_conf = TLS_BASE;
 	ctx->rx_conf = TLS_BASE;
+	ctx->sk_proto = sk->sk_prot;
 	update_sk_prot(sk, ctx);
 out:
 	return rc;


^ permalink raw reply related

* Re: [PATCH 00/15] Netfilter/IPVS updates for net-next
From: David Miller @ 2019-07-08 19:14 UTC (permalink / raw)
  To: pablo; +Cc: netfilter-devel, netdev
In-Reply-To: <20190708103237.28061-1-pablo@netfilter.org>

From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Mon,  8 Jul 2019 12:32:22 +0200

> The following patchset contains Netfilter/IPVS updates for net-next:
 ...
> You can pull these changes from:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next.git

Pulled, thanks.

^ permalink raw reply

* [bpf PATCH v2 1/6] tls: remove close callback sock unlock/lock and flush_sync
From: John Fastabend @ 2019-07-08 19:13 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf
In-Reply-To: <156261310104.31108.4569969631798277807.stgit@ubuntu3-kvm1>

The tls close() callback currently drops the sock lock, makes a
cancel_delayed_work_sync() call, and then relocks the sock. The
same pattern is used again to call strp_done().

By restructuring the code we can avoid droping lock and then
reclaiming it. To simplify this we do the following,

 tls_sk_proto_close
  set_bit(CLOSING)
  set_bit(SCHEDULE)
  cancel_delay_work_sync() <- cancel workqueue
  lock_sock(sk)
  ...
  release_sock(sk)
  strp_done()

Setting the CLOSING bit prevents the SCHEDULE bit from being
cleared by any workqueue items e.g. if one happens to be
scheduled and run between when we set SCHEDULE bit and cancel
work. Then because SCHEDULE bit is set now no new work will
be scheduled.

Then strp_done() is called after the sk work is completed.
Any outstanding work is sync'd and finally ctx is free'd.

Tested with net selftests and bpf selftests.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 include/net/tls.h  |    4 ++--
 net/tls/tls_main.c |   56 ++++++++++++++++++++++++++--------------------------
 net/tls/tls_sw.c   |   38 ++++++++++++++++++++++-------------
 3 files changed, 54 insertions(+), 44 deletions(-)

diff --git a/include/net/tls.h b/include/net/tls.h
index 63e473420b00..0a3540a6304d 100644
--- a/include/net/tls.h
+++ b/include/net/tls.h
@@ -107,9 +107,7 @@ struct tls_device {
 enum {
 	TLS_BASE,
 	TLS_SW,
-#ifdef CONFIG_TLS_DEVICE
 	TLS_HW,
-#endif
 	TLS_HW_RECORD,
 	TLS_NUM_CONFIG,
 };
@@ -162,6 +160,7 @@ struct tls_sw_context_tx {
 	int async_capable;
 
 #define BIT_TX_SCHEDULED	0
+#define BIT_TX_CLOSING		1
 	unsigned long tx_bitmask;
 };
 
@@ -361,6 +360,7 @@ void tls_sw_close(struct sock *sk, long timeout);
 void tls_sw_free_resources_tx(struct sock *sk);
 void tls_sw_free_resources_rx(struct sock *sk);
 void tls_sw_release_resources_rx(struct sock *sk);
+void tls_sw_release_strp_rx(struct tls_context *tls_ctx);
 int tls_sw_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 		   int nonblock, int flags, int *addr_len);
 bool tls_sw_stream_read(const struct sock *sk);
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index fc81ae18cc44..d63c3583d2f7 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -261,24 +261,9 @@ static void tls_ctx_free(struct tls_context *ctx)
 	kfree(ctx);
 }
 
-static void tls_sk_proto_close(struct sock *sk, long timeout)
+static void tls_sk_proto_cleanup(struct sock *sk,
+				 struct tls_context *ctx, long timeo)
 {
-	struct tls_context *ctx = tls_get_ctx(sk);
-	long timeo = sock_sndtimeo(sk, 0);
-	void (*sk_proto_close)(struct sock *sk, long timeout);
-	bool free_ctx = false;
-
-	lock_sock(sk);
-	sk_proto_close = ctx->sk_proto_close;
-
-	if (ctx->tx_conf == TLS_HW_RECORD && ctx->rx_conf == TLS_HW_RECORD)
-		goto skip_tx_cleanup;
-
-	if (ctx->tx_conf == TLS_BASE && ctx->rx_conf == TLS_BASE) {
-		free_ctx = true;
-		goto skip_tx_cleanup;
-	}
-
 	if (!tls_complete_pending_work(sk, ctx, 0, &timeo))
 		tls_handle_open_record(sk, 0);
 
@@ -299,22 +284,37 @@ static void tls_sk_proto_close(struct sock *sk, long timeout)
 #ifdef CONFIG_TLS_DEVICE
 	if (ctx->rx_conf == TLS_HW)
 		tls_device_offload_cleanup_rx(sk);
-
-	if (ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW) {
-#else
-	{
 #endif
-		tls_ctx_free(ctx);
-		ctx = NULL;
-	}
+}
+
+static void tls_sk_proto_close(struct sock *sk, long timeout)
+{
+	struct tls_context *ctx = tls_get_ctx(sk);
+	long timeo = sock_sndtimeo(sk, 0);
+	void (*sk_proto_close)(struct sock *sk, long timeout);
+
+	if (ctx->tx_conf == TLS_SW)
+		tls_sw_cancel_work_tx(ctx);
+
+	lock_sock(sk);
+	sk_proto_close = ctx->sk_proto_close;
+
+	if (ctx->tx_conf == TLS_HW_RECORD && ctx->rx_conf == TLS_HW_RECORD)
+		goto skip_tx_cleanup;
+
+	if (ctx->tx_conf == TLS_BASE && ctx->rx_conf == TLS_BASE)
+		goto skip_tx_cleanup;
+
+	tls_sk_proto_cleanup(sk, ctx, timeo);
 
 skip_tx_cleanup:
 	release_sock(sk);
+	if (ctx->rx_conf == TLS_SW)
+		tls_sw_release_strp_rx(ctx);
 	sk_proto_close(sk, timeout);
-	/* free ctx for TLS_HW_RECORD, used by tcp_set_state
-	 * for sk->sk_prot->unhash [tls_hw_unhash]
-	 */
-	if (free_ctx)
+
+	if (ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW &&
+	    ctx->tx_conf != TLS_HW_RECORD && ctx->rx_conf != TLS_HW_RECORD)
 		tls_ctx_free(ctx);
 }
 
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index db585964b52b..3b01cd72ca6c 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -2053,6 +2053,15 @@ static void tls_data_ready(struct sock *sk)
 	}
 }
 
+void tls_sw_cancel_work_tx(struct tls_context *tls_ctx)
+{
+	struct tls_sw_context_tx *ctx = tls_sw_ctx_tx(tls_ctx);
+
+	set_bit(BIT_TX_CLOSING, &ctx->tx_bitmask);
+	set_bit(BIT_TX_SCHEDULED, &ctx->tx_bitmask);
+	cancel_delayed_work_sync(&ctx->tx_work.work);
+}
+
 void tls_sw_free_resources_tx(struct sock *sk)
 {
 	struct tls_context *tls_ctx = tls_get_ctx(sk);
@@ -2064,11 +2073,6 @@ void tls_sw_free_resources_tx(struct sock *sk)
 	if (atomic_read(&ctx->encrypt_pending))
 		crypto_wait_req(-EINPROGRESS, &ctx->async_wait);
 
-	release_sock(sk);
-	cancel_delayed_work_sync(&ctx->tx_work.work);
-	lock_sock(sk);
-
-	/* Tx whatever records we can transmit and abandon the rest */
 	tls_tx_records(sk, -1);
 
 	/* Free up un-sent records in tx_list. First, free
@@ -2112,22 +2116,22 @@ void tls_sw_release_resources_rx(struct sock *sk)
 		write_lock_bh(&sk->sk_callback_lock);
 		sk->sk_data_ready = ctx->saved_data_ready;
 		write_unlock_bh(&sk->sk_callback_lock);
-		release_sock(sk);
-		strp_done(&ctx->strp);
-		lock_sock(sk);
 	}
 }
 
-void tls_sw_free_resources_rx(struct sock *sk)
+void tls_sw_release_strp_rx(struct tls_context *tls_ctx)
 {
-	struct tls_context *tls_ctx = tls_get_ctx(sk);
 	struct tls_sw_context_rx *ctx = tls_sw_ctx_rx(tls_ctx);
 
-	tls_sw_release_resources_rx(sk);
-
+	strp_done(&ctx->strp);
 	kfree(ctx);
 }
 
+void tls_sw_free_resources_rx(struct sock *sk)
+{
+	tls_sw_release_resources_rx(sk);
+}
+
 /* The work handler to transmitt the encrypted records in tx_list */
 static void tx_work_handler(struct work_struct *work)
 {
@@ -2136,11 +2140,17 @@ static void tx_work_handler(struct work_struct *work)
 					       struct tx_work, work);
 	struct sock *sk = tx_work->sk;
 	struct tls_context *tls_ctx = tls_get_ctx(sk);
-	struct tls_sw_context_tx *ctx = tls_sw_ctx_tx(tls_ctx);
+	struct tls_sw_context_tx *ctx;
 
-	if (!test_and_clear_bit(BIT_TX_SCHEDULED, &ctx->tx_bitmask))
+	if (unlikely(!tls_ctx))
+		return;
+
+	ctx = tls_sw_ctx_tx(tls_ctx);
+	if (test_bit(BIT_TX_CLOSING, &ctx->tx_bitmask))
 		return;
 
+	if (!test_and_clear_bit(BIT_TX_SCHEDULED, &ctx->tx_bitmask))
+		return;
 	lock_sock(sk);
 	tls_tx_records(sk, -1);
 	release_sock(sk);


^ permalink raw reply related

* [bpf PATCH v2 0/6] bpf: sockmap/tls fixes
From: John Fastabend @ 2019-07-08 19:13 UTC (permalink / raw)
  To: jakub.kicinski, ast, daniel; +Cc: netdev, edumazet, john.fastabend, bpf

Resolve a series of splats discovered by syzbot and an unhash
TLS issue noted by Eric Dumazet.

The main issues revolved around interaction between TLS and
sockmap tear down. TLS and sockmap could both reset sk->prot
ops creating a condition where a close or unhash op could be
called forever. A rare race condition resulting from a missing
rcu sync operation was causing a use after free. Then on the
TLS side dropping the sock lock and re-acquiring it during the
close op could hang. Finally, sockmap must be deployed before
tls for current stack assumptions to be met. This is enforced
now. A feature series can enable it.

To fix this first refactor TLS code so the lock is held for the
entire teardown operation. Then add an unhash callback to ensure
TLS can not transition from ESTABLISHED to LISTEN state. This
transition is a similar bug to the one found and fixed previously
in sockmap. Then apply three fixes to sockmap to fix up races
on tear down around map free and close. Finally, if sockmap
is destroyed before TLS we add a new ULP op update to inform
the TLS stack it should not call sockmap ops. This last one
appears to be the most commonly found issue from syzbot.

---

John Fastabend (6):
      tls: remove close callback sock unlock/lock and flush_sync
      bpf: tls fix transition through disconnect with close
      bpf: sockmap, sock_map_delete needs to use xchg
      bpf: sockmap, synchronize_rcu before free'ing map
      bpf: sockmap, only create entry if ulp is not already enabled
      bpf: sockmap/tls, close can race with map free


 include/linux/skmsg.h |    8 +++
 include/net/tcp.h     |    3 +
 include/net/tls.h     |   10 +++-
 net/core/skmsg.c      |    4 +
 net/core/sock_map.c   |   19 +++++--
 net/ipv4/tcp_ulp.c    |   13 +++++
 net/tls/tls_main.c    |  135 ++++++++++++++++++++++++++++++++++++++-----------
 net/tls/tls_sw.c      |   38 +++++++++-----
 8 files changed, 176 insertions(+), 54 deletions(-)

--
Signature

^ permalink raw reply

* Re: [PATCH net-next v2 8/8] net: mscc: PTP Hardware Clock (PHC) support
From: Jakub Kicinski @ 2019-07-08 19:06 UTC (permalink / raw)
  To: Antoine Tenart
  Cc: davem, richardcochran, alexandre.belloni, UNGLinuxDriver, ralf,
	paul.burton, jhogan, netdev, linux-mips, thomas.petazzoni,
	allan.nielsen
In-Reply-To: <20190708084809.GB2932@kwain>

On Mon, 8 Jul 2019 10:48:09 +0200, Antoine Tenart wrote:
> > > +	/* Commit back the result & save it */
> > > +	memcpy(&ocelot->hwtstamp_config, &cfg, sizeof(cfg));
> > > +	mutex_unlock(&ocelot->ptp_lock);
> > > +
> > > +	return copy_to_user(ifr->ifr_data, &cfg, sizeof(cfg)) ? -EFAULT : 0;
> > > +}
> > >  
> > > +static int ocelot_get_ts_info(struct net_device *dev,
> > > +			      struct ethtool_ts_info *info)
> > > +{
> > > +	struct ocelot_port *ocelot_port = netdev_priv(dev);
> > > +	struct ocelot *ocelot = ocelot_port->ocelot;
> > > +	int ret;
> > > +
> > > +	if (!ocelot->ptp)
> > > +		return -EOPNOTSUPP;  
> > 
> > Hmm.. why does software timestamping depend on PTP?  
> 
> Because it depends on the "PTP" register bank (and the "PTP" interrupt)
> being described and available. This is why I named the flag 'ptp', but
> it could be named 'timestamp' or 'ts' as well.

Right, but software timestamps are done by calling skb_tx_timestamp(skb)
in the driver, no need for HW support there (software RX timestamp is
handled by the stack).

^ permalink raw reply

* Re: [PATCH] net: netsec: Sync dma for device on buffer allocation
From: David Miller @ 2019-07-08 19:00 UTC (permalink / raw)
  To: ilias.apalodimas; +Cc: netdev, jaswinder.singh
In-Reply-To: <1562570741-25108-1-git-send-email-ilias.apalodimas@linaro.org>

From: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Date: Mon,  8 Jul 2019 10:25:41 +0300

> cd1973a9215a ("net: netsec: Sync dma for device on buffer allocation")
> was merged on it's v1 instead of the v3.
> Merge the proper patch version
> 
> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>

Applied to net-next, thanks.

^ permalink raw reply

* Re: [PATCH net-next 4/4] bnxt_en: add page_pool support
From: Andy Gospodarek @ 2019-07-08 18:49 UTC (permalink / raw)
  To: Michael Chan, Ilias Apalodimas; +Cc: davem, netdev, hawk, ast
In-Reply-To: <1562398578-26020-5-git-send-email-michael.chan@broadcom.com>

On Sat, Jul 06, 2019 at 03:36:18AM -0400, Michael Chan wrote:
> From: Andy Gospodarek <gospo@broadcom.com>
> 
> This removes contention over page allocation for XDP_REDIRECT actions by
> adding page_pool support per queue for the driver.  The performance for
> XDP_REDIRECT actions scales linearly with the number of cores performing
> redirect actions when using the page pools instead of the standard page
> allocator.
> 
> Signed-off-by: Andy Gospodarek <gospo@broadcom.com>
> Signed-off-by: Michael Chan <michael.chan@broadcom.com>
> ---
>  drivers/net/ethernet/broadcom/Kconfig         |  1 +
>  drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 40 +++++++++++++++++++++++----
>  drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  3 ++
>  drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c |  3 +-
>  4 files changed, 41 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> index d8f0846..b6777e5 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
[...]
> @@ -2530,12 +2555,17 @@ static int bnxt_alloc_rx_rings(struct bnxt *bp)
>  
>  		ring = &rxr->rx_ring_struct;
>  
> +		rc = bnxt_alloc_rx_page_pool(bp, rxr);
> +		if (rc)
> +			return rc;
> +
>  		rc = xdp_rxq_info_reg(&rxr->xdp_rxq, bp->dev, i);
>  		if (rc < 0)
>  			return rc;
>  
>  		rc = xdp_rxq_info_reg_mem_model(&rxr->xdp_rxq,
> -						MEM_TYPE_PAGE_SHARED, NULL);
> +						MEM_TYPE_PAGE_POOL,
> +						rxr->page_pool);
>  		if (rc) {
>  			xdp_rxq_info_unreg(&rxr->xdp_rxq);
>  			return rc;

I think we want to amend and the chunk above to be:

@@ -2530,14 +2557,24 @@ static int bnxt_alloc_rx_rings(struct bnxt *bp)
 
                ring = &rxr->rx_ring_struct;
 
+               rc = bnxt_alloc_rx_page_pool(bp, rxr);
+               if (rc)
+                       return rc;
+
                rc = xdp_rxq_info_reg(&rxr->xdp_rxq, bp->dev, i);
-               if (rc < 0)
+               if (rc < 0) {
+                       page_pool_free(rxr->page_pool);
+                       rxr->page_pool = NULL;
                        return rc;
+               }
 
                rc = xdp_rxq_info_reg_mem_model(&rxr->xdp_rxq,
-                                               MEM_TYPE_PAGE_SHARED, NULL);
+                                               MEM_TYPE_PAGE_POOL,
+                                               rxr->page_pool);
                if (rc) {
                        xdp_rxq_info_unreg(&rxr->xdp_rxq);
+                       page_pool_free(rxr->page_pool);
+                       rxr->page_pool = NULL;
                        return rc;
                }
 

That should take care of the freeing of the page_pool that is allocated
but there is a failure during xdp_rxq_info_reg() or
xdp_rxq_info_reg_mem_model().

I agree that we do not need to call page_pool_free in the normal
shutdown case.

^ permalink raw reply

* Re: use exact allocation for dma coherent memory
From: Christoph Hellwig @ 2019-07-08 18:43 UTC (permalink / raw)
  To: Arend Van Spriel
  Cc: Christoph Hellwig, Maarten Lankhorst, Maxime Ripard, Sean Paul,
	David Airlie, Daniel Vetter, Jani Nikula, Joonas Lahtinen,
	Rodrigo Vivi, Ian Abbott, H Hartley Sweeten, devel, linux-s390,
	Intel Linux Wireless, linux-rdma, netdev, intel-gfx,
	linux-wireless, linux-kernel, dri-devel, linux-mm, iommu,
	moderated list:ARM PORT, linux-media
In-Reply-To: <74eb9d99-6aa6-d1ad-e66d-6cc9c496b2f3@broadcom.com>

On Tue, Jul 02, 2019 at 11:48:44AM +0200, Arend Van Spriel wrote:
> You made me look ;-) Actually not touching my drivers so I'm off the hook. 
> However, I was wondering if drivers could know so I decided to look into 
> the DMA-API.txt documentation which currently states:
>
> """
> The flag parameter (dma_alloc_coherent() only) allows the caller to
> specify the ``GFP_`` flags (see kmalloc()) for the allocation (the
> implementation may choose to ignore flags that affect the location of
> the returned memory, like GFP_DMA).
> """
>
> I do expect you are going to change that description as well now that you 
> are going to issue a warning on __GFP_COMP. Maybe include that in patch 
> 15/16 where you introduce that warning.

Yes, that description needs an updated, even without this series.
I'll make sure it is more clear.

^ permalink raw reply

* Re: [PATCH] r8169: add enable_aspm parameter
From: Heiner Kallweit @ 2019-07-08 18:27 UTC (permalink / raw)
  To: AceLan Kao, Realtek linux nic maintainers, David S. Miller,
	netdev, linux-kernel
In-Reply-To: <20190708063751.16234-1-acelan.kao@canonical.com>

On 08.07.2019 08:37, AceLan Kao wrote:
> We have many commits in the driver which enable and then disable ASPM
> function over and over again.
>    commit b75bb8a5b755 ("r8169: disable ASPM again")
>    commit 0866cd15029b ("r8169: enable ASPM on RTL8106E")
>    commit 94235460f9ea ("r8169: Align ASPM/CLKREQ setting function with vendor driver")
>    commit aa1e7d2c31ef ("r8169: enable ASPM on RTL8168E-VL")
>    commit f37658da21aa ("r8169: align ASPM entry latency setting with vendor driver")
>    commit a99790bf5c7f ("r8169: Reinstate ASPM Support")
>    commit 671646c151d4 ("r8169: Don't disable ASPM in the driver")
>    commit 4521e1a94279 ("Revert "r8169: enable internal ASPM and clock request settings".")
>    commit d64ec841517a ("r8169: enable internal ASPM and clock request settings")
> 
> This function is very important for production, and if we can't come out
> a solution to make both happy, I'd suggest we add a parameter in the
> driver to toggle it.
> 
The usage of a module parameter to control ASPM is discouraged.
There have been more such attempts in the past that have been declined.

Pending with the PCI maintainers is a series adding ASPM control
via sysfs, see here: https://www.spinics.net/lists/linux-pci/msg83228.html

Also more details than just stating "it's important for production"
would have been appreciated in the commit message, e.g. which
power-savings you can achieve with ASPM on which systems.

^ permalink raw reply


This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox