* [PATCH v5 12/15] octeontx2-af: Set RVU PFs to CGX LMACs mapping
From: sunil.kovvuri @ 2018-10-01 11:36 UTC (permalink / raw)
To: netdev, davem
Cc: arnd, linux-soc, Linu Cherian, Geetha sowjanya, Sunil Goutham
In-Reply-To: <1538393791-26251-1-git-send-email-sunil.kovvuri@gmail.com>
From: Linu Cherian <lcherian@marvell.com>
Each of the enabled CGX LMAC is considered a physical
interface and RVU PFs are mapped to these. VFs of these
SRIOV PFs will be virtual interfaces and share CGX LMAC
along with PF.
This mapping info will be used later on for Rx/Tx pkt steering.
Signed-off-by: Linu Cherian <lcherian@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
---
drivers/net/ethernet/marvell/octeontx2/af/Makefile | 2 +-
drivers/net/ethernet/marvell/octeontx2/af/cgx.c | 59 ++++++++++++++++++++++
drivers/net/ethernet/marvell/octeontx2/af/cgx.h | 15 +++++-
drivers/net/ethernet/marvell/octeontx2/af/rvu.c | 4 ++
drivers/net/ethernet/marvell/octeontx2/af/rvu.h | 12 +++++
5 files changed, 89 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/Makefile b/drivers/net/ethernet/marvell/octeontx2/af/Makefile
index 8646421..eaac264 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/Makefile
+++ b/drivers/net/ethernet/marvell/octeontx2/af/Makefile
@@ -7,4 +7,4 @@ obj-$(CONFIG_OCTEONTX2_MBOX) += octeontx2_mbox.o
obj-$(CONFIG_OCTEONTX2_AF) += octeontx2_af.o
octeontx2_mbox-y := mbox.o
-octeontx2_af-y := cgx.o rvu.o
+octeontx2_af-y := cgx.o rvu.o rvu_cgx.o
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cgx.c b/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
index 17ff3081..f7a48bd 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
+++ b/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
@@ -28,8 +28,12 @@ struct cgx {
void __iomem *reg_base;
struct pci_dev *pdev;
u8 cgx_id;
+ u8 lmac_count;
+ struct list_head cgx_list;
};
+static LIST_HEAD(cgx_list);
+
/* Supported devices */
static const struct pci_device_id cgx_id_table[] = {
{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVID_OCTEONTX2_CGX) },
@@ -38,6 +42,53 @@ static const struct pci_device_id cgx_id_table[] = {
MODULE_DEVICE_TABLE(pci, cgx_id_table);
+static u64 cgx_read(struct cgx *cgx, u64 lmac, u64 offset)
+{
+ return readq(cgx->reg_base + (lmac << 18) + offset);
+}
+
+int cgx_get_cgx_cnt(void)
+{
+ struct cgx *cgx_dev;
+ int count = 0;
+
+ list_for_each_entry(cgx_dev, &cgx_list, cgx_list)
+ count++;
+
+ return count;
+}
+EXPORT_SYMBOL(cgx_get_cgx_cnt);
+
+int cgx_get_lmac_cnt(void *cgxd)
+{
+ struct cgx *cgx = cgxd;
+
+ if (!cgx)
+ return -ENODEV;
+
+ return cgx->lmac_count;
+}
+EXPORT_SYMBOL(cgx_get_lmac_cnt);
+
+void *cgx_get_pdata(int cgx_id)
+{
+ struct cgx *cgx_dev;
+
+ list_for_each_entry(cgx_dev, &cgx_list, cgx_list) {
+ if (cgx_dev->cgx_id == cgx_id)
+ return cgx_dev;
+ }
+ return NULL;
+}
+EXPORT_SYMBOL(cgx_get_pdata);
+
+static void cgx_lmac_init(struct cgx *cgx)
+{
+ cgx->lmac_count = cgx_read(cgx, 0, CGXX_CMRX_RX_LMACS) & 0x7;
+ if (cgx->lmac_count > MAX_LMAC_PER_CGX)
+ cgx->lmac_count = MAX_LMAC_PER_CGX;
+}
+
static int cgx_probe(struct pci_dev *pdev, const struct pci_device_id *id)
{
int err;
@@ -72,9 +123,14 @@ static int cgx_probe(struct pci_dev *pdev, const struct pci_device_id *id)
goto err_release_regions;
}
+ list_add(&cgx->cgx_list, &cgx_list);
+ cgx->cgx_id = cgx_get_cgx_cnt() - 1;
+ cgx_lmac_init(cgx);
+
return 0;
err_release_regions:
+ list_del(&cgx->cgx_list);
pci_release_regions(pdev);
err_disable_device:
pci_disable_device(pdev);
@@ -84,6 +140,9 @@ static int cgx_probe(struct pci_dev *pdev, const struct pci_device_id *id)
static void cgx_remove(struct pci_dev *pdev)
{
+ struct cgx *cgx = pci_get_drvdata(pdev);
+
+ list_del(&cgx->cgx_list);
pci_release_regions(pdev);
pci_disable_device(pdev);
pci_set_drvdata(pdev, NULL);
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cgx.h b/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
index a7d4b39..acdc16e 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
+++ b/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
@@ -12,11 +12,22 @@
#define CGX_H
/* PCI device IDs */
-#define PCI_DEVID_OCTEONTX2_CGX 0xA059
+#define PCI_DEVID_OCTEONTX2_CGX 0xA059
/* PCI BAR nos */
-#define PCI_CFG_REG_BAR_NUM 0
+#define PCI_CFG_REG_BAR_NUM 0
+
+#define MAX_CGX 3
+#define MAX_LMAC_PER_CGX 4
+#define CGX_OFFSET(x) ((x) * MAX_LMAC_PER_CGX)
+
+/* Registers */
+#define CGXX_CMRX_RX_ID_MAP 0x060
+#define CGXX_CMRX_RX_LMACS 0x128
extern struct pci_driver cgx_driver;
+int cgx_get_cgx_cnt(void);
+int cgx_get_lmac_cnt(void *cgxd);
+void *cgx_get_pdata(int cgx_id);
#endif /* CGX_H */
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
index 2e9b252..b363d19 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
@@ -1558,6 +1558,10 @@ static int rvu_probe(struct pci_dev *pdev, const struct pci_device_id *id)
if (err)
goto err_hwsetup;
+ err = rvu_cgx_probe(rvu);
+ if (err)
+ goto err_mbox;
+
err = rvu_register_interrupts(rvu);
if (err)
goto err_mbox;
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
index 92c2022..385f597 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
@@ -100,6 +100,16 @@ struct rvu {
char *irq_name;
bool *irq_allocated;
dma_addr_t msix_base_iova;
+
+ /* CGX */
+#define PF_CGXMAP_BASE 1 /* PF 0 is reserved for RVU PF */
+ u8 cgx_mapped_pfs;
+ u8 cgx_cnt; /* available cgx ports */
+ u8 *pf2cgxlmac_map; /* pf to cgx_lmac map */
+ u16 *cgxlmac2pf_map; /* bitmap of mapped pfs for
+ * every cgx lmac port
+ */
+ void **cgx_idmap; /* cgx id to cgx data map table */
};
static inline void rvu_write64(struct rvu *rvu, u64 block, u64 offset, u64 val)
@@ -138,4 +148,6 @@ int rvu_get_lf(struct rvu *rvu, struct rvu_block *block, u16 pcifunc, u16 slot);
int rvu_get_blkaddr(struct rvu *rvu, int blktype, u16 pcifunc);
int rvu_poll_reg(struct rvu *rvu, u64 block, u64 offset, u64 mask, bool zero);
+/* CGX APIs */
+int rvu_cgx_probe(struct rvu *rvu);
#endif /* RVU_H */
--
2.7.4
^ permalink raw reply related
* [PATCH v5 13/15] octeontx2-af: Add support for CGX link management
From: sunil.kovvuri @ 2018-10-01 11:36 UTC (permalink / raw)
To: netdev, davem; +Cc: arnd, linux-soc, Linu Cherian, Nithya Mani, Sunil Goutham
In-Reply-To: <1538393791-26251-1-git-send-email-sunil.kovvuri@gmail.com>
From: Linu Cherian <lcherian@marvell.com>
CGX LMAC initialization, link status polling etc is done
by low level secure firmware. For link management this patch
adds a interface or communication mechanism between firmware
and this kernel CGX driver.
- Firmware interface specification is defined in cgx_fw_if.h.
- Support to send/receive commands/events to/form firmware.
- events/commands implemented
* link up
* link down
* reading firmware version
Signed-off-by: Linu Cherian <lcherian@marvell.com>
Signed-off-by: Nithya Mani <nmani@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
---
drivers/net/ethernet/marvell/octeontx2/af/cgx.c | 364 ++++++++++++++++++++-
drivers/net/ethernet/marvell/octeontx2/af/cgx.h | 32 ++
.../net/ethernet/marvell/octeontx2/af/cgx_fw_if.h | 292 +++++++++++++++++
.../net/ethernet/marvell/octeontx2/af/rvu_cgx.c | 97 ++++++
4 files changed, 781 insertions(+), 4 deletions(-)
create mode 100644 drivers/net/ethernet/marvell/octeontx2/af/cgx_fw_if.h
create mode 100644 drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cgx.c b/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
index f7a48bd..7e24706 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
+++ b/drivers/net/ethernet/marvell/octeontx2/af/cgx.c
@@ -24,16 +24,43 @@
#define DRV_NAME "octeontx2-cgx"
#define DRV_STRING "Marvell OcteonTX2 CGX/MAC Driver"
+/**
+ * struct lmac
+ * @wq_cmd_cmplt: waitq to keep the process blocked until cmd completion
+ * @cmd_lock: Lock to serialize the command interface
+ * @resp: command response
+ * @event_cb: callback for linkchange events
+ * @cmd_pend: flag set before new command is started
+ * flag cleared after command response is received
+ * @cgx: parent cgx port
+ * @lmac_id: lmac port id
+ * @name: lmac port name
+ */
+struct lmac {
+ wait_queue_head_t wq_cmd_cmplt;
+ struct mutex cmd_lock;
+ struct cgx_evt_sts resp;
+ struct cgx_event_cb event_cb;
+ bool cmd_pend;
+ struct cgx *cgx;
+ u8 lmac_id;
+ char *name;
+};
+
struct cgx {
void __iomem *reg_base;
struct pci_dev *pdev;
u8 cgx_id;
u8 lmac_count;
+ struct lmac *lmac_idmap[MAX_LMAC_PER_CGX];
struct list_head cgx_list;
};
static LIST_HEAD(cgx_list);
+/* CGX PHY management internal APIs */
+static int cgx_fwi_link_change(struct cgx *cgx, int lmac_id, bool en);
+
/* Supported devices */
static const struct pci_device_id cgx_id_table[] = {
{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVID_OCTEONTX2_CGX) },
@@ -42,11 +69,24 @@ static const struct pci_device_id cgx_id_table[] = {
MODULE_DEVICE_TABLE(pci, cgx_id_table);
+static void cgx_write(struct cgx *cgx, u64 lmac, u64 offset, u64 val)
+{
+ writeq(val, cgx->reg_base + (lmac << 18) + offset);
+}
+
static u64 cgx_read(struct cgx *cgx, u64 lmac, u64 offset)
{
return readq(cgx->reg_base + (lmac << 18) + offset);
}
+static inline struct lmac *lmac_pdata(u8 lmac_id, struct cgx *cgx)
+{
+ if (!cgx || lmac_id >= MAX_LMAC_PER_CGX)
+ return NULL;
+
+ return cgx->lmac_idmap[lmac_id];
+}
+
int cgx_get_cgx_cnt(void)
{
struct cgx *cgx_dev;
@@ -82,18 +122,318 @@ void *cgx_get_pdata(int cgx_id)
}
EXPORT_SYMBOL(cgx_get_pdata);
-static void cgx_lmac_init(struct cgx *cgx)
+/* CGX Firmware interface low level support */
+static int cgx_fwi_cmd_send(struct cgx_cmd *cmd, struct cgx_evt_sts *rsp,
+ struct lmac *lmac)
+{
+ struct cgx *cgx = lmac->cgx;
+ union cgx_cmdreg creg;
+ union cgx_evtreg ereg;
+ struct device *dev;
+ int err = 0;
+
+ /* Ensure no other command is in progress */
+ err = mutex_lock_interruptible(&lmac->cmd_lock);
+ if (err)
+ return err;
+
+ /* Ensure command register is free */
+ creg.val = cgx_read(cgx, lmac->lmac_id, CGX_COMMAND_REG);
+ if (creg.cmd.own != CGX_CMD_OWN_NS) {
+ err = -EBUSY;
+ goto unlock;
+ }
+
+ /* Update ownership in command request */
+ cmd->own = CGX_CMD_OWN_FIRMWARE;
+
+ /* Mark this lmac as pending, before we start */
+ lmac->cmd_pend = true;
+
+ /* Start command in hardware */
+ creg.cmd = *cmd;
+ cgx_write(cgx, lmac->lmac_id, CGX_COMMAND_REG, creg.val);
+ creg.val = cgx_read(cgx, lmac->lmac_id, CGX_COMMAND_REG);
+
+ /* Ensure command is completed without errors */
+ if (!wait_event_timeout(lmac->wq_cmd_cmplt, !lmac->cmd_pend,
+ msecs_to_jiffies(CGX_CMD_TIMEOUT))) {
+ dev = &cgx->pdev->dev;
+ ereg.val = cgx_read(cgx, lmac->lmac_id, CGX_EVENT_REG);
+ if (ereg.val) {
+ dev_err(dev, "cgx port %d:%d: No event for response\n",
+ cgx->cgx_id, lmac->lmac_id);
+ /* copy event */
+ lmac->resp = ereg.evt_sts;
+ } else {
+ dev_err(dev, "cgx port %d:%d cmd timeout\n",
+ cgx->cgx_id, lmac->lmac_id);
+ err = -EIO;
+ goto unlock;
+ }
+ }
+
+ /* we have a valid command response */
+ smp_rmb(); /* Ensure the latest updates are visible */
+ *rsp = lmac->resp;
+
+unlock:
+ mutex_unlock(&lmac->cmd_lock);
+
+ return err;
+}
+
+static inline int cgx_fwi_cmd_generic(struct cgx_cmd *req,
+ struct cgx_evt_sts *rsp,
+ struct cgx *cgx, int lmac_id)
+{
+ struct lmac *lmac;
+ int err;
+
+ lmac = lmac_pdata(lmac_id, cgx);
+ if (!lmac)
+ return -ENODEV;
+
+ err = cgx_fwi_cmd_send(req, rsp, lmac);
+
+ /* Check for valid response */
+ if (!err) {
+ if (rsp->stat == CGX_STAT_FAIL)
+ return -EIO;
+ else
+ return 0;
+ }
+
+ return err;
+}
+
+/* Hardware event handlers */
+static inline void cgx_link_change_handler(struct cgx_lnk_sts *lstat,
+ struct lmac *lmac)
+{
+ struct cgx *cgx = lmac->cgx;
+ struct cgx_link_event event;
+ struct device *dev = &cgx->pdev->dev;
+
+ event.lstat = *lstat;
+ event.cgx_id = cgx->cgx_id;
+ event.lmac_id = lmac->lmac_id;
+
+ if (!lmac->event_cb.notify_link_chg) {
+ dev_dbg(dev, "cgx port %d:%d Link change handler null",
+ cgx->cgx_id, lmac->lmac_id);
+ if (lstat->err_type != CGX_ERR_NONE) {
+ dev_err(dev, "cgx port %d:%d Link error %d\n",
+ cgx->cgx_id, lmac->lmac_id, lstat->err_type);
+ }
+ dev_info(dev, "cgx port %d:%d Link status %s, speed %x\n",
+ cgx->cgx_id, lmac->lmac_id,
+ lstat->link_up ? "UP" : "DOWN", lstat->speed);
+ return;
+ }
+
+ if (lmac->event_cb.notify_link_chg(&event, lmac->event_cb.data))
+ dev_err(dev, "event notification failure\n");
+}
+
+static inline bool cgx_cmdresp_is_linkevent(struct cgx_evt_sts *rsp)
+{
+ if (rsp->id == CGX_CMD_LINK_BRING_UP ||
+ rsp->id == CGX_CMD_LINK_BRING_DOWN)
+ return true;
+ else
+ return false;
+}
+
+static inline bool cgx_event_is_linkevent(struct cgx_evt_sts *evt)
+{
+ if (evt->id == CGX_EVT_LINK_CHANGE)
+ return true;
+ else
+ return false;
+}
+
+static irqreturn_t cgx_fwi_event_handler(int irq, void *data)
+{
+ struct lmac *lmac = data;
+ struct cgx *cgx = lmac->cgx;
+ struct cgx_evt_sts event;
+ union cgx_evtreg ereg;
+ struct device *dev;
+
+ ereg.val = cgx_read(cgx, lmac->lmac_id, CGX_EVENT_REG);
+ if (!ereg.evt_sts.ack)
+ return IRQ_NONE;
+
+ dev = &cgx->pdev->dev;
+
+ event = ereg.evt_sts;
+
+ switch (event.evt_type) {
+ case CGX_EVT_CMD_RESP:
+ /* Copy the response. Since only one command is active at a
+ * time, there is no way a response can get overwritten
+ */
+ lmac->resp = event;
+ /* Ensure response is updated before thread context starts */
+ smp_wmb();
+
+ /* There wont be separate events for link change initiated from
+ * software; Hence report the command responses as events
+ */
+ if (cgx_cmdresp_is_linkevent(&event))
+ cgx_link_change_handler(&ereg.link_sts, lmac);
+
+ /* Release thread waiting for completion */
+ lmac->cmd_pend = false;
+ wake_up_interruptible(&lmac->wq_cmd_cmplt);
+ break;
+ case CGX_EVT_ASYNC:
+ if (cgx_event_is_linkevent(&event))
+ cgx_link_change_handler(&ereg.link_sts, lmac);
+ break;
+ default:
+ dev_err(dev, "cgx port %d:%d Unknown event received\n",
+ cgx->cgx_id, lmac->lmac_id);
+ }
+
+ /* Any new event or command response will be posted by firmware
+ * only after the current status is acked.
+ * Ack the interrupt register as well.
+ */
+ cgx_write(lmac->cgx, lmac->lmac_id, CGX_EVENT_REG, 0);
+ cgx_write(lmac->cgx, lmac->lmac_id, CGXX_CMRX_INT, FW_CGX_INT);
+
+ return IRQ_HANDLED;
+}
+
+/* APIs for PHY management using CGX firmware interface */
+
+/* callback registration for hardware events like link change */
+int cgx_lmac_evh_register(struct cgx_event_cb *cb, void *cgxd, int lmac_id)
{
+ struct lmac *lmac;
+ struct cgx *cgx = cgxd;
+
+ lmac = lmac_pdata(lmac_id, cgx);
+ if (!lmac)
+ return -ENODEV;
+
+ lmac->event_cb = *cb;
+
+ return 0;
+}
+EXPORT_SYMBOL(cgx_lmac_evh_register);
+
+static int cgx_fwi_link_change(struct cgx *cgx, int lmac_id, bool enable)
+{
+ struct cgx_cmd req = { 0 };
+ struct cgx_evt_sts rsp;
+
+ if (enable)
+ req.id = CGX_CMD_LINK_BRING_UP;
+ else
+ req.id = CGX_CMD_LINK_BRING_DOWN;
+
+ return cgx_fwi_cmd_generic(&req, &rsp, cgx, lmac_id);
+}
+EXPORT_SYMBOL(cgx_fwi_link_change);
+
+static inline int cgx_fwi_read_version(struct cgx_ver_s *ver, struct cgx *cgx)
+{
+ struct cgx_cmd req = { 0 };
+ union cgx_evtreg event;
+ int err;
+
+ req.id = CGX_CMD_GET_FW_VER;
+
+ err = cgx_fwi_cmd_generic(&req, &event.evt_sts, cgx, 0);
+ if (!err)
+ *ver = event.ver;
+
+ return err;
+}
+
+static int cgx_lmac_verify_fwi_version(struct cgx *cgx)
+{
+ struct cgx_ver_s ver;
+ struct device *dev = &cgx->pdev->dev;
+ int err;
+
+ if (!cgx->lmac_count)
+ return 0;
+
+ err = cgx_fwi_read_version(&ver, cgx);
+ dev_dbg(dev, "Firmware command interface version = %d.%d\n",
+ ver.major_ver, ver.minor_ver);
+ if (err || ver.major_ver != CGX_FIRMWARE_MAJOR_VER ||
+ ver.minor_ver != CGX_FIRMWARE_MINOR_VER)
+ return -EIO;
+ else
+ return 0;
+}
+
+static int cgx_lmac_init(struct cgx *cgx)
+{
+ struct lmac *lmac;
+ int i, err;
+
cgx->lmac_count = cgx_read(cgx, 0, CGXX_CMRX_RX_LMACS) & 0x7;
if (cgx->lmac_count > MAX_LMAC_PER_CGX)
cgx->lmac_count = MAX_LMAC_PER_CGX;
+
+ for (i = 0; i < cgx->lmac_count; i++) {
+ lmac = kcalloc(1, sizeof(struct lmac), GFP_KERNEL);
+ if (!lmac)
+ return -ENOMEM;
+ lmac->name = kcalloc(1, sizeof("cgx_fwi_xxx_yyy"), GFP_KERNEL);
+ if (!lmac->name)
+ return -ENOMEM;
+ sprintf(lmac->name, "cgx_fwi_%d_%d", cgx->cgx_id, i);
+ lmac->lmac_id = i;
+ lmac->cgx = cgx;
+ init_waitqueue_head(&lmac->wq_cmd_cmplt);
+ mutex_init(&lmac->cmd_lock);
+ err = request_irq(pci_irq_vector(cgx->pdev,
+ CGX_LMAC_FWI + i * 9),
+ cgx_fwi_event_handler, 0, lmac->name, lmac);
+ if (err)
+ return err;
+
+ /* Enable interrupt */
+ cgx_write(cgx, lmac->lmac_id, CGXX_CMRX_INT_ENA_W1S,
+ FW_CGX_INT);
+
+ /* Add reference */
+ cgx->lmac_idmap[i] = lmac;
+ }
+
+ return cgx_lmac_verify_fwi_version(cgx);
+}
+
+static int cgx_lmac_exit(struct cgx *cgx)
+{
+ struct lmac *lmac;
+ int i;
+
+ /* Free all lmac related resources */
+ for (i = 0; i < cgx->lmac_count; i++) {
+ lmac = cgx->lmac_idmap[i];
+ if (!lmac)
+ continue;
+ free_irq(pci_irq_vector(cgx->pdev, CGX_LMAC_FWI + i * 9), lmac);
+ kfree(lmac->name);
+ kfree(lmac);
+ }
+
+ return 0;
}
static int cgx_probe(struct pci_dev *pdev, const struct pci_device_id *id)
{
- int err;
struct device *dev = &pdev->dev;
struct cgx *cgx;
+ int err, nvec;
cgx = devm_kzalloc(dev, sizeof(*cgx), GFP_KERNEL);
if (!cgx)
@@ -123,14 +463,28 @@ static int cgx_probe(struct pci_dev *pdev, const struct pci_device_id *id)
goto err_release_regions;
}
+ nvec = CGX_NVEC;
+ err = pci_alloc_irq_vectors(pdev, nvec, nvec, PCI_IRQ_MSIX);
+ if (err < 0 || err != nvec) {
+ dev_err(dev, "Request for %d msix vectors failed, err %d\n",
+ nvec, err);
+ goto err_release_regions;
+ }
+
list_add(&cgx->cgx_list, &cgx_list);
cgx->cgx_id = cgx_get_cgx_cnt() - 1;
- cgx_lmac_init(cgx);
+
+ err = cgx_lmac_init(cgx);
+ if (err)
+ goto err_release_lmac;
+
return 0;
-err_release_regions:
+err_release_lmac:
+ cgx_lmac_exit(cgx);
list_del(&cgx->cgx_list);
+err_release_regions:
pci_release_regions(pdev);
err_disable_device:
pci_disable_device(pdev);
@@ -142,7 +496,9 @@ static void cgx_remove(struct pci_dev *pdev)
{
struct cgx *cgx = pci_get_drvdata(pdev);
+ cgx_lmac_exit(cgx);
list_del(&cgx->cgx_list);
+ pci_free_irq_vectors(pdev);
pci_release_regions(pdev);
pci_disable_device(pdev);
pci_set_drvdata(pdev, NULL);
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cgx.h b/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
index acdc16e..a2a7a6d 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
+++ b/drivers/net/ethernet/marvell/octeontx2/af/cgx.h
@@ -11,6 +11,8 @@
#ifndef CGX_H
#define CGX_H
+#include "cgx_fw_if.h"
+
/* PCI device IDs */
#define PCI_DEVID_OCTEONTX2_CGX 0xA059
@@ -22,12 +24,42 @@
#define CGX_OFFSET(x) ((x) * MAX_LMAC_PER_CGX)
/* Registers */
+#define CGXX_CMRX_INT 0x040
+#define FW_CGX_INT BIT_ULL(1)
+#define CGXX_CMRX_INT_ENA_W1S 0x058
#define CGXX_CMRX_RX_ID_MAP 0x060
#define CGXX_CMRX_RX_LMACS 0x128
+#define CGXX_SCRATCH0_REG 0x1050
+#define CGXX_SCRATCH1_REG 0x1058
+#define CGX_CONST 0x2000
+
+#define CGX_COMMAND_REG CGXX_SCRATCH1_REG
+#define CGX_EVENT_REG CGXX_SCRATCH0_REG
+#define CGX_CMD_TIMEOUT 2200 /* msecs */
+
+#define CGX_NVEC 37
+#define CGX_LMAC_FWI 0
+
+struct cgx_link_event {
+ struct cgx_lnk_sts lstat;
+ u8 cgx_id;
+ u8 lmac_id;
+};
+
+/**
+ * struct cgx_event_cb
+ * @notify_link_chg: callback for link change notification
+ * @data: data passed to callback function
+ */
+struct cgx_event_cb {
+ int (*notify_link_chg)(struct cgx_link_event *event, void *data);
+ void *data;
+};
extern struct pci_driver cgx_driver;
int cgx_get_cgx_cnt(void);
int cgx_get_lmac_cnt(void *cgxd);
void *cgx_get_pdata(int cgx_id);
+int cgx_lmac_evh_register(struct cgx_event_cb *cb, void *cgxd, int lmac_id);
#endif /* CGX_H */
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cgx_fw_if.h b/drivers/net/ethernet/marvell/octeontx2/af/cgx_fw_if.h
new file mode 100644
index 0000000..6a03e7d
--- /dev/null
+++ b/drivers/net/ethernet/marvell/octeontx2/af/cgx_fw_if.h
@@ -0,0 +1,292 @@
+/* SPDX-License-Identifier: GPL-2.0
+ * Marvell OcteonTx2 CGX driver
+ *
+ * Copyright (C) 2018 Marvell International Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef __CGX_FW_INTF_H__
+#define __CGX_FW_INTF_H__
+
+#define CGX_FIRMWARE_MAJOR_VER 1
+#define CGX_FIRMWARE_MINOR_VER 0
+
+#define CGX_EVENT_ACK 1UL
+
+/* CGX error types. set for cmd response status as CGX_STAT_FAIL */
+enum cgx_error_type {
+ CGX_ERR_NONE,
+ CGX_ERR_LMAC_NOT_ENABLED,
+ CGX_ERR_LMAC_MODE_INVALID,
+ CGX_ERR_REQUEST_ID_INVALID,
+ CGX_ERR_PREV_ACK_NOT_CLEAR,
+ CGX_ERR_PHY_LINK_DOWN,
+ CGX_ERR_PCS_RESET_FAIL,
+ CGX_ERR_AN_CPT_FAIL,
+ CGX_ERR_TX_NOT_IDLE,
+ CGX_ERR_RX_NOT_IDLE,
+ CGX_ERR_SPUX_BR_BLKLOCK_FAIL,
+ CGX_ERR_SPUX_RX_ALIGN_FAIL,
+ CGX_ERR_SPUX_TX_FAULT,
+ CGX_ERR_SPUX_RX_FAULT,
+ CGX_ERR_SPUX_RESET_FAIL,
+ CGX_ERR_SPUX_AN_RESET_FAIL,
+ CGX_ERR_SPUX_USX_AN_RESET_FAIL,
+ CGX_ERR_SMUX_RX_LINK_NOT_OK,
+ CGX_ERR_PCS_RECV_LINK_FAIL,
+ CGX_ERR_TRAINING_FAIL,
+ CGX_ERR_RX_EQU_FAIL,
+ CGX_ERR_SPUX_BER_FAIL,
+ CGX_ERR_SPUX_RSFEC_ALGN_FAIL, /* = 22 */
+};
+
+/* LINK speed types */
+enum cgx_link_speed {
+ CGX_LINK_NONE,
+ CGX_LINK_10M,
+ CGX_LINK_100M,
+ CGX_LINK_1G,
+ CGX_LINK_2HG,
+ CGX_LINK_5G,
+ CGX_LINK_10G,
+ CGX_LINK_20G,
+ CGX_LINK_25G,
+ CGX_LINK_40G,
+ CGX_LINK_50G,
+ CGX_LINK_100G,
+ CGX_LINK_SPEED_MAX,
+};
+
+/* REQUEST ID types. Input to firmware */
+enum cgx_cmd_id {
+ CGX_CMD_NONE,
+ CGX_CMD_GET_FW_VER,
+ CGX_CMD_GET_MAC_ADDR,
+ CGX_CMD_SET_MTU,
+ CGX_CMD_GET_LINK_STS, /* optional to user */
+ CGX_CMD_LINK_BRING_UP,
+ CGX_CMD_LINK_BRING_DOWN,
+ CGX_CMD_INTERNAL_LBK,
+ CGX_CMD_EXTERNAL_LBK,
+ CGX_CMD_HIGIG,
+ CGX_CMD_LINK_STATE_CHANGE,
+ CGX_CMD_MODE_CHANGE, /* hot plug support */
+ CGX_CMD_INTF_SHUTDOWN,
+ CGX_CMD_IRQ_ENABLE,
+ CGX_CMD_IRQ_DISABLE,
+};
+
+/* async event ids */
+enum cgx_evt_id {
+ CGX_EVT_NONE,
+ CGX_EVT_LINK_CHANGE,
+};
+
+/* event types - cause of interrupt */
+enum cgx_evt_type {
+ CGX_EVT_ASYNC,
+ CGX_EVT_CMD_RESP
+};
+
+enum cgx_stat {
+ CGX_STAT_SUCCESS,
+ CGX_STAT_FAIL
+};
+
+enum cgx_cmd_own {
+ CGX_CMD_OWN_NS,
+ CGX_CMD_OWN_FIRMWARE,
+};
+
+/* scratchx(0) CSR used for ATF->non-secure SW communication.
+ * This acts as the status register
+ * Provides details on command ack/status, link status, error details
+ */
+struct cgx_evt_sts {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved:55;
+ uint64_t id:6;
+ uint64_t stat:1;
+ uint64_t evt_type:1;
+ uint64_t ack:1;
+#else
+ uint64_t ack:1;
+ uint64_t evt_type:1; /* cgx_evt_type */
+ uint64_t stat:1; /* cgx_stat */
+ uint64_t id:6; /* cgx_evt_id/cgx_cmd_id */
+ uint64_t reserved:55;
+#endif
+};
+
+/* Response to command IDs with command status as CGX_STAT_FAIL
+ *
+ * Not applicable for commands :
+ * CGX_CMD_LINK_BRING_UP/DOWN/CGX_EVT_LINK_CHANGE
+ * check struct cgx_lnk_sts comments
+ */
+struct cgx_err_sts_s {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:35;
+ uint64_t type:10;
+ uint64_t reserved1:9;
+#else
+ uint64_t reserved1:9;
+ uint64_t type:10; /* cgx_error_type */
+ uint64_t reserved2:35;
+#endif
+};
+
+/* Response to cmd ID as CGX_CMD_GET_FW_VER with cmd status as
+ * CGX_STAT_SUCCESS
+ */
+struct cgx_ver_s {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:47;
+ uint64_t minor_ver:4;
+ uint64_t major_ver:4;
+ uint64_t reserved1:9;
+#else
+ uint64_t reserved1:9;
+ uint64_t major_ver:4;
+ uint64_t minor_ver:4;
+ uint64_t reserved2:47;
+#endif
+};
+
+/* Response to cmd ID as CGX_CMD_GET_MAC_ADDR with cmd status as
+ * CGX_STAT_SUCCESS
+ */
+struct cgx_mac_addr_s {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:7;
+ uint64_t addr_5:8;
+ uint64_t addr_4:8;
+ uint64_t addr_3:8;
+ uint64_t addr_2:8;
+ uint64_t addr_1:8;
+ uint64_t addr_0:8;
+ uint64_t reserved1:9;
+#else
+ uint64_t reserved1:9;
+ uint64_t local_mac_addr:48;
+ uint64_t reserved2:7;
+#endif
+};
+
+/* Response to cmd ID - CGX_CMD_LINK_BRING_UP/DOWN, event ID CGX_EVT_LINK_CHANGE
+ * status can be either CGX_STAT_FAIL or CGX_STAT_SUCCESS
+ *
+ * In case of CGX_STAT_FAIL, it indicates CGX configuration failed
+ * when processing link up/down/change command.
+ * Both err_type and current link status will be updated
+ *
+ * In case of CGX_STAT_SUCCESS, err_type will be CGX_ERR_NONE and current
+ * link status will be updated
+ */
+struct cgx_lnk_sts {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:39;
+ uint64_t err_type:10;
+ uint64_t speed:4;
+ uint64_t full_duplex:1;
+ uint64_t link_up:1;
+ uint64_t reserved1:9;
+#else
+ uint64_t reserved1:9;
+ uint64_t link_up:1;
+ uint64_t full_duplex:1;
+ uint64_t speed:4; /* cgx_link_speed */
+ uint64_t err_type:10;
+ uint64_t reserved2:39;
+#endif
+};
+
+union cgx_evtreg {
+ u64 val;
+ struct cgx_evt_sts evt_sts; /* common for all commands/events */
+ struct cgx_lnk_sts link_sts; /* response to LINK_BRINGUP/DOWN/CHANGE */
+ struct cgx_ver_s ver; /* response to CGX_CMD_GET_FW_VER */
+ struct cgx_mac_addr_s mac_addr; /* response to CGX_CMD_GET_MAC_ADDR */
+ struct cgx_err_sts_s err; /* response if evt_status = CMD_FAIL */
+};
+
+/* scratchx(1) CSR used for non-secure SW->ATF communication
+ * This CSR acts as a command register
+ */
+struct cgx_cmd {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:56;
+ uint64_t id:6; /* cgx_request_id */
+ uint64_t own:2; /* cgx_csr_own */
+#else
+ uint64_t own:2; /* cgx_csr_own */
+ uint64_t id:6; /* cgx_request_id */
+ uint64_t reserved2:56;
+#endif
+};
+
+/* Any command using enable/disable as an argument need
+ * to pass the option via this structure.
+ * Ex: Loopback, HiGig...
+ */
+struct cgx_ctl_args {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:55;
+ uint64_t enable:1;
+ uint64_t reserved1:8;
+#else
+ uint64_t reserved1:8;
+ uint64_t enable:1;
+ uint64_t reserved2:55;
+#endif
+};
+
+/* command argument to be passed for cmd ID - CGX_CMD_SET_MTU */
+struct cgx_mtu_args {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:40;
+ uint64_t size:16;
+ uint64_t reserved1:8;
+#else
+ uint64_t reserved1:8;
+ uint64_t size:16;
+ uint64_t reserved2:40;
+#endif
+};
+
+/* command argument to be passed for cmd ID - CGX_CMD_LINK_CHANGE */
+struct cgx_link_change_args {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ uint64_t reserved2:50;
+ uint64_t speed:4;
+ uint64_t full_duplex:1;
+ uint64_t link_up:1;
+ uint64_t reserved1:8;
+#else
+ uint64_t reserved1:8;
+ uint64_t link_up:1;
+ uint64_t full_duplex:1;
+ uint64_t speed:4; /* cgx_link_speed */
+ uint64_t reserved2:50;
+#endif
+};
+
+struct cgx_irq_cfg {
+ uint64_t reserved1:8;
+ uint64_t irq_phys:32;
+ uint64_t reserved2:24;
+};
+
+union cgx_cmdreg {
+ u64 val;
+ struct cgx_cmd cmd;
+ struct cgx_ctl_args cmd_args;
+ struct cgx_mtu_args mtu_size;
+ struct cgx_irq_cfg irq_cfg; /* Input to CGX_CMD_IRQ_ENABLE */
+ struct cgx_link_change_args lnk_args;/* Input to CGX_CMD_LINK_CHANGE */
+};
+
+#endif /* __CGX_FW_INTF_H__ */
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
new file mode 100644
index 0000000..bf81507
--- /dev/null
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
@@ -0,0 +1,97 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Marvell OcteonTx2 RVU Admin Function driver
+ *
+ * Copyright (C) 2018 Marvell International Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/types.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+
+#include "rvu.h"
+#include "cgx.h"
+
+static inline u8 cgxlmac_id_to_bmap(u8 cgx_id, u8 lmac_id)
+{
+ return ((cgx_id & 0xF) << 4) | (lmac_id & 0xF);
+}
+
+static void *rvu_cgx_pdata(u8 cgx_id, struct rvu *rvu)
+{
+ if (cgx_id >= rvu->cgx_cnt)
+ return NULL;
+
+ return rvu->cgx_idmap[cgx_id];
+}
+
+static int rvu_map_cgx_lmac_pf(struct rvu *rvu)
+{
+ int cgx_cnt = rvu->cgx_cnt;
+ int cgx, lmac_cnt, lmac;
+ int pf = PF_CGXMAP_BASE;
+ int size;
+
+ if (!cgx_cnt)
+ return 0;
+
+ if (cgx_cnt > 0xF || MAX_LMAC_PER_CGX > 0xF)
+ return -EINVAL;
+
+ /* Alloc map table
+ * An additional entry is required since PF id starts from 1 and
+ * hence entry at offset 0 is invalid.
+ */
+ size = (cgx_cnt * MAX_LMAC_PER_CGX + 1) * sizeof(u8);
+ rvu->pf2cgxlmac_map = devm_kzalloc(rvu->dev, size, GFP_KERNEL);
+ if (!rvu->pf2cgxlmac_map)
+ return -ENOMEM;
+
+ /* Initialize offset 0 with an invalid cgx and lmac id */
+ rvu->pf2cgxlmac_map[0] = 0xFF;
+
+ /* Reverse map table */
+ rvu->cgxlmac2pf_map = devm_kzalloc(rvu->dev,
+ cgx_cnt * MAX_LMAC_PER_CGX * sizeof(u16),
+ GFP_KERNEL);
+ if (!rvu->cgxlmac2pf_map)
+ return -ENOMEM;
+
+ rvu->cgx_mapped_pfs = 0;
+ for (cgx = 0; cgx < cgx_cnt; cgx++) {
+ lmac_cnt = cgx_get_lmac_cnt(rvu_cgx_pdata(cgx, rvu));
+ for (lmac = 0; lmac < lmac_cnt; lmac++, pf++) {
+ rvu->pf2cgxlmac_map[pf] = cgxlmac_id_to_bmap(cgx, lmac);
+ rvu->cgxlmac2pf_map[CGX_OFFSET(cgx) + lmac] = 1 << pf;
+ rvu->cgx_mapped_pfs++;
+ }
+ }
+ return 0;
+}
+
+int rvu_cgx_probe(struct rvu *rvu)
+{
+ int i;
+
+ /* find available cgx ports */
+ rvu->cgx_cnt = cgx_get_cgx_cnt();
+ if (!rvu->cgx_cnt) {
+ dev_info(rvu->dev, "No CGX devices found!\n");
+ return -ENODEV;
+ }
+
+ rvu->cgx_idmap = devm_kzalloc(rvu->dev, rvu->cgx_cnt * sizeof(void *),
+ GFP_KERNEL);
+ if (!rvu->cgx_idmap)
+ return -ENOMEM;
+
+ /* Initialize the cgxdata table */
+ for (i = 0; i < rvu->cgx_cnt; i++)
+ rvu->cgx_idmap[i] = cgx_get_pdata(i);
+
+ /* Map CGX LMAC interfaces to RVU PFs */
+ return rvu_map_cgx_lmac_pf(rvu);
+}
--
2.7.4
^ permalink raw reply related
* [PATCH v5 14/15] octeontx2-af: Register for CGX lmac events
From: sunil.kovvuri @ 2018-10-01 11:36 UTC (permalink / raw)
To: netdev, davem; +Cc: arnd, linux-soc, Linu Cherian, Sunil Goutham
In-Reply-To: <1538393791-26251-1-git-send-email-sunil.kovvuri@gmail.com>
From: Linu Cherian <lcherian@marvell.com>
Added support in RVU AF driver to register for
CGX LMAC link status change events from firmware
and managing them. Processing part will be added
in followup patches.
- Introduced eventqueue for posting events from cgx lmac.
Queueing mechanism will ensure that events can be posted
and firmware can be acked immediately and hence event
reception and processing are decoupled.
- Events gets added to the queue by notification callback.
Notification callback is expected to be atomic, since it
is called from interrupt context.
- Events are dequeued and processed in a worker thread.
Signed-off-by: Linu Cherian <lcherian@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
---
drivers/net/ethernet/marvell/octeontx2/af/rvu.c | 6 +-
drivers/net/ethernet/marvell/octeontx2/af/rvu.h | 5 +
.../net/ethernet/marvell/octeontx2/af/rvu_cgx.c | 101 ++++++++++++++++++++-
3 files changed, 108 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
index b363d19..adc7fc6 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu.c
@@ -1564,10 +1564,11 @@ static int rvu_probe(struct pci_dev *pdev, const struct pci_device_id *id)
err = rvu_register_interrupts(rvu);
if (err)
- goto err_mbox;
+ goto err_cgx;
return 0;
-
+err_cgx:
+ rvu_cgx_wq_destroy(rvu);
err_mbox:
rvu_mbox_destroy(rvu);
err_hwsetup:
@@ -1589,6 +1590,7 @@ static void rvu_remove(struct pci_dev *pdev)
struct rvu *rvu = pci_get_drvdata(pdev);
rvu_unregister_interrupts(rvu);
+ rvu_cgx_wq_destroy(rvu);
rvu_mbox_destroy(rvu);
rvu_reset_all_blocks(rvu);
rvu_free_hw_resources(rvu);
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
index 385f597..d169fa9 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h
@@ -110,6 +110,10 @@ struct rvu {
* every cgx lmac port
*/
void **cgx_idmap; /* cgx id to cgx data map table */
+ struct work_struct cgx_evh_work;
+ struct workqueue_struct *cgx_evh_wq;
+ spinlock_t cgx_evq_lock; /* cgx event queue lock */
+ struct list_head cgx_evq_head; /* cgx event queue head */
};
static inline void rvu_write64(struct rvu *rvu, u64 block, u64 offset, u64 val)
@@ -150,4 +154,5 @@ int rvu_poll_reg(struct rvu *rvu, u64 block, u64 offset, u64 mask, bool zero);
/* CGX APIs */
int rvu_cgx_probe(struct rvu *rvu);
+void rvu_cgx_wq_destroy(struct rvu *rvu);
#endif /* RVU_H */
diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
index bf81507..2359806e 100644
--- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
+++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_cgx.c
@@ -15,6 +15,11 @@
#include "rvu.h"
#include "cgx.h"
+struct cgx_evq_entry {
+ struct list_head evq_node;
+ struct cgx_link_event link_event;
+};
+
static inline u8 cgxlmac_id_to_bmap(u8 cgx_id, u8 lmac_id)
{
return ((cgx_id & 0xF) << 4) | (lmac_id & 0xF);
@@ -72,9 +77,95 @@ static int rvu_map_cgx_lmac_pf(struct rvu *rvu)
return 0;
}
+/* This is called from interrupt context and is expected to be atomic */
+static int cgx_lmac_postevent(struct cgx_link_event *event, void *data)
+{
+ struct rvu *rvu = data;
+ struct cgx_evq_entry *qentry;
+
+ /* post event to the event queue */
+ qentry = kmalloc(sizeof(*qentry), GFP_ATOMIC);
+ if (!qentry)
+ return -ENOMEM;
+ qentry->link_event = *event;
+ spin_lock(&rvu->cgx_evq_lock);
+ list_add_tail(&qentry->evq_node, &rvu->cgx_evq_head);
+ spin_unlock(&rvu->cgx_evq_lock);
+
+ /* start worker to process the events */
+ queue_work(rvu->cgx_evh_wq, &rvu->cgx_evh_work);
+
+ return 0;
+}
+
+static void cgx_evhandler_task(struct work_struct *work)
+{
+ struct rvu *rvu = container_of(work, struct rvu, cgx_evh_work);
+ struct cgx_evq_entry *qentry;
+ struct cgx_link_event *event;
+ unsigned long flags;
+
+ do {
+ /* Dequeue an event */
+ spin_lock_irqsave(&rvu->cgx_evq_lock, flags);
+ qentry = list_first_entry_or_null(&rvu->cgx_evq_head,
+ struct cgx_evq_entry,
+ evq_node);
+ if (qentry)
+ list_del(&qentry->evq_node);
+ spin_unlock_irqrestore(&rvu->cgx_evq_lock, flags);
+ if (!qentry)
+ break; /* nothing more to process */
+
+ event = &qentry->link_event;
+
+ /* Do nothing for now */
+ kfree(qentry);
+ } while (1);
+}
+
+static void cgx_lmac_event_handler_init(struct rvu *rvu)
+{
+ struct cgx_event_cb cb;
+ int cgx, lmac, err;
+ void *cgxd;
+
+ spin_lock_init(&rvu->cgx_evq_lock);
+ INIT_LIST_HEAD(&rvu->cgx_evq_head);
+ INIT_WORK(&rvu->cgx_evh_work, cgx_evhandler_task);
+ rvu->cgx_evh_wq = alloc_workqueue("rvu_evh_wq", 0, 0);
+ if (!rvu->cgx_evh_wq) {
+ dev_err(rvu->dev, "alloc workqueue failed");
+ return;
+ }
+
+ cb.notify_link_chg = cgx_lmac_postevent; /* link change call back */
+ cb.data = rvu;
+
+ for (cgx = 0; cgx < rvu->cgx_cnt; cgx++) {
+ cgxd = rvu_cgx_pdata(cgx, rvu);
+ for (lmac = 0; lmac < cgx_get_lmac_cnt(cgxd); lmac++) {
+ err = cgx_lmac_evh_register(&cb, cgxd, lmac);
+ if (err)
+ dev_err(rvu->dev,
+ "%d:%d handler register failed\n",
+ cgx, lmac);
+ }
+ }
+}
+
+void rvu_cgx_wq_destroy(struct rvu *rvu)
+{
+ if (rvu->cgx_evh_wq) {
+ flush_workqueue(rvu->cgx_evh_wq);
+ destroy_workqueue(rvu->cgx_evh_wq);
+ rvu->cgx_evh_wq = NULL;
+ }
+}
+
int rvu_cgx_probe(struct rvu *rvu)
{
- int i;
+ int i, err;
/* find available cgx ports */
rvu->cgx_cnt = cgx_get_cgx_cnt();
@@ -93,5 +184,11 @@ int rvu_cgx_probe(struct rvu *rvu)
rvu->cgx_idmap[i] = cgx_get_pdata(i);
/* Map CGX LMAC interfaces to RVU PFs */
- return rvu_map_cgx_lmac_pf(rvu);
+ err = rvu_map_cgx_lmac_pf(rvu);
+ if (err)
+ return err;
+
+ /* Register for CGX events */
+ cgx_lmac_event_handler_init(rvu);
+ return 0;
}
--
2.7.4
^ permalink raw reply related
* [PATCH v5 15/15] MAINTAINERS: Add entry for Marvell OcteonTX2 Admin Function driver
From: sunil.kovvuri @ 2018-10-01 11:36 UTC (permalink / raw)
To: netdev, davem; +Cc: arnd, linux-soc, Sunil Goutham
In-Reply-To: <1538393791-26251-1-git-send-email-sunil.kovvuri@gmail.com>
From: Sunil Goutham <sgoutham@marvell.com>
Added maintainers entry for Marvell OcteonTX2 SOC's RVU
admin function driver.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
---
MAINTAINERS | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 15565de..cf12fa3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8843,6 +8843,15 @@ S: Supported
F: drivers/mmc/host/sdhci-xenon*
F: Documentation/devicetree/bindings/mmc/marvell,xenon-sdhci.txt
+MARVELL OCTEONTX2 RVU ADMIN FUNCTION DRIVER
+M: Sunil Goutham <sgoutham@marvell.com>
+M: Linu Cherian <lcherian@marvell.com>
+M: Geetha sowjanya <gakula@marvell.com>
+M: Jerin Jacob <jerinj@marvell.com>
+L: netdev@vger.kernel.org
+S: Maintained
+F: drivers/net/ethernet/marvell/octeontx2/af
+
MATROX FRAMEBUFFER DRIVER
L: linux-fbdev@vger.kernel.org
S: Orphan
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 0/7] Support of Flow Director in HNS3 Ethernet Driver for HiP08 Rev2 SoC
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm
This patch-set adds the support of FD(Flow Director) in the HNS3 PF driver
for HiP08 Rev2(0x21) SoC of Hisilicon. FD can be used in filtering the flows
and deciding to drop the flow or forward it to paricular queue.
Configuration consists of rules with input keys and actions. The rules are
stored in TCAM.
Jian Shen (7):
net: hns3: Add flow director initialization
net: hns3: Add input key and action config support for flow director
net: hns3: Add support for rule add/delete for flow director
net: hns3: Add support for rule query of flow director
net: hns3: Add reset handle for flow director
net: hns3: Remove all flow director rules when unload hns3 driver
net: hns3: Add support for enable/disable flow director
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 19 +
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 55 +-
drivers/net/ethernet/hisilicon/hns3/hns3_enet.h | 5 +
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 39 +-
.../net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 77 ++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 1303 ++++++++++++++++++++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 232 ++++
7 files changed, 1721 insertions(+), 9 deletions(-)
--
2.7.4
^ permalink raw reply
* [PATCH net-next 4/7] net: hns3: Add support for rule query of flow director
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm, Jian Shen
In-Reply-To: <20181001114647.15504-1-salil.mehta@huawei.com>
From: Jian Shen <shenjian15@huawei.com>
This patch adds support for querying rule number and rule details
by ethtool commands.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 6 +
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 25 ++-
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 239 +++++++++++++++++++++
3 files changed, 264 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index 262bb736..fac84d8 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -420,6 +420,12 @@ struct hnae3_ae_ops {
struct ethtool_rxnfc *cmd);
int (*del_fd_entry)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd);
+ int (*get_fd_rule_cnt)(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd);
+ int (*get_fd_rule_info)(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd);
+ int (*get_fd_all_rules)(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd, u32 *rule_locs);
};
struct hnae3_dcb_ops {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 59cbf78..7d79a07 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -699,20 +699,33 @@ static int hns3_get_rxnfc(struct net_device *netdev,
{
struct hnae3_handle *h = hns3_get_handle(netdev);
- if (!h->ae_algo || !h->ae_algo->ops || !h->ae_algo->ops->get_rss_tuple)
+ if (!h->ae_algo || !h->ae_algo->ops)
return -EOPNOTSUPP;
switch (cmd->cmd) {
case ETHTOOL_GRXRINGS:
- cmd->data = h->kinfo.rss_size;
- break;
+ cmd->data = h->kinfo.num_tqps;
+ return 0;
case ETHTOOL_GRXFH:
- return h->ae_algo->ops->get_rss_tuple(h, cmd);
+ if (h->ae_algo->ops->get_rss_tuple)
+ return h->ae_algo->ops->get_rss_tuple(h, cmd);
+ return -EOPNOTSUPP;
+ case ETHTOOL_GRXCLSRLCNT:
+ if (h->ae_algo->ops->get_fd_rule_cnt)
+ return h->ae_algo->ops->get_fd_rule_cnt(h, cmd);
+ return -EOPNOTSUPP;
+ case ETHTOOL_GRXCLSRULE:
+ if (h->ae_algo->ops->get_fd_rule_info)
+ return h->ae_algo->ops->get_fd_rule_info(h, cmd);
+ return -EOPNOTSUPP;
+ case ETHTOOL_GRXCLSRLALL:
+ if (h->ae_algo->ops->get_fd_all_rules)
+ return h->ae_algo->ops->get_fd_all_rules(h, cmd,
+ rule_locs);
+ return -EOPNOTSUPP;
default:
return -EOPNOTSUPP;
}
-
- return 0;
}
static int hns3_change_all_ring_bd_num(struct hns3_nic_priv *priv,
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 43f5caa..876c7ca 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -4299,6 +4299,242 @@ static int hclge_del_fd_entry(struct hnae3_handle *handle,
false);
}
+static int hclge_get_fd_rule_cnt(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ cmd->rule_cnt = hdev->hclge_fd_rule_num;
+ cmd->data = hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1];
+
+ return 0;
+}
+
+static int hclge_get_fd_rule_info(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_fd_rule *rule = NULL;
+ struct hclge_dev *hdev = vport->back;
+ struct ethtool_rx_flow_spec *fs;
+ struct hlist_node *node2;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ fs = (struct ethtool_rx_flow_spec *)&cmd->fs;
+
+ hlist_for_each_entry_safe(rule, node2, &hdev->fd_rule_list, rule_node) {
+ if (rule->location >= fs->location)
+ break;
+ }
+
+ if (!rule || fs->location != rule->location)
+ return -ENOENT;
+
+ fs->flow_type = rule->flow_type;
+ switch (fs->flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) {
+ case SCTP_V4_FLOW:
+ case TCP_V4_FLOW:
+ case UDP_V4_FLOW:
+ fs->h_u.tcp_ip4_spec.ip4src =
+ cpu_to_be32(rule->tuples.src_ip[3]);
+ fs->m_u.tcp_ip4_spec.ip4src =
+ rule->unused_tuple & BIT(INNER_SRC_IP) ?
+ 0 : cpu_to_be32(rule->tuples_mask.src_ip[3]);
+
+ fs->h_u.tcp_ip4_spec.ip4dst =
+ cpu_to_be32(rule->tuples.dst_ip[3]);
+ fs->m_u.tcp_ip4_spec.ip4dst =
+ rule->unused_tuple & BIT(INNER_DST_IP) ?
+ 0 : cpu_to_be32(rule->tuples_mask.dst_ip[3]);
+
+ fs->h_u.tcp_ip4_spec.psrc = cpu_to_be16(rule->tuples.src_port);
+ fs->m_u.tcp_ip4_spec.psrc =
+ rule->unused_tuple & BIT(INNER_SRC_PORT) ?
+ 0 : cpu_to_be16(rule->tuples_mask.src_port);
+
+ fs->h_u.tcp_ip4_spec.pdst = cpu_to_be16(rule->tuples.dst_port);
+ fs->m_u.tcp_ip4_spec.pdst =
+ rule->unused_tuple & BIT(INNER_DST_PORT) ?
+ 0 : cpu_to_be16(rule->tuples_mask.dst_port);
+
+ fs->h_u.tcp_ip4_spec.tos = rule->tuples.ip_tos;
+ fs->m_u.tcp_ip4_spec.tos =
+ rule->unused_tuple & BIT(INNER_IP_TOS) ?
+ 0 : rule->tuples_mask.ip_tos;
+
+ break;
+ case IP_USER_FLOW:
+ fs->h_u.usr_ip4_spec.ip4src =
+ cpu_to_be32(rule->tuples.src_ip[3]);
+ fs->m_u.tcp_ip4_spec.ip4src =
+ rule->unused_tuple & BIT(INNER_SRC_IP) ?
+ 0 : cpu_to_be32(rule->tuples_mask.src_ip[3]);
+
+ fs->h_u.usr_ip4_spec.ip4dst =
+ cpu_to_be32(rule->tuples.dst_ip[3]);
+ fs->m_u.usr_ip4_spec.ip4dst =
+ rule->unused_tuple & BIT(INNER_DST_IP) ?
+ 0 : cpu_to_be32(rule->tuples_mask.dst_ip[3]);
+
+ fs->h_u.usr_ip4_spec.tos = rule->tuples.ip_tos;
+ fs->m_u.usr_ip4_spec.tos =
+ rule->unused_tuple & BIT(INNER_IP_TOS) ?
+ 0 : rule->tuples_mask.ip_tos;
+
+ fs->h_u.usr_ip4_spec.proto = rule->tuples.ip_proto;
+ fs->m_u.usr_ip4_spec.proto =
+ rule->unused_tuple & BIT(INNER_IP_PROTO) ?
+ 0 : rule->tuples_mask.ip_proto;
+
+ fs->h_u.usr_ip4_spec.ip_ver = ETH_RX_NFC_IP4;
+
+ break;
+ case SCTP_V6_FLOW:
+ case TCP_V6_FLOW:
+ case UDP_V6_FLOW:
+ cpu_to_be32_array(fs->h_u.tcp_ip6_spec.ip6src,
+ rule->tuples.src_ip, 4);
+ if (rule->unused_tuple & BIT(INNER_SRC_IP))
+ memset(fs->m_u.tcp_ip6_spec.ip6src, 0, sizeof(int) * 4);
+ else
+ cpu_to_be32_array(fs->m_u.tcp_ip6_spec.ip6src,
+ rule->tuples_mask.src_ip, 4);
+
+ cpu_to_be32_array(fs->h_u.tcp_ip6_spec.ip6dst,
+ rule->tuples.dst_ip, 4);
+ if (rule->unused_tuple & BIT(INNER_DST_IP))
+ memset(fs->m_u.tcp_ip6_spec.ip6dst, 0, sizeof(int) * 4);
+ else
+ cpu_to_be32_array(fs->m_u.tcp_ip6_spec.ip6dst,
+ rule->tuples_mask.dst_ip, 4);
+
+ fs->h_u.tcp_ip6_spec.psrc = cpu_to_be16(rule->tuples.src_port);
+ fs->m_u.tcp_ip6_spec.psrc =
+ rule->unused_tuple & BIT(INNER_SRC_PORT) ?
+ 0 : cpu_to_be16(rule->tuples_mask.src_port);
+
+ fs->h_u.tcp_ip6_spec.pdst = cpu_to_be16(rule->tuples.dst_port);
+ fs->m_u.tcp_ip6_spec.pdst =
+ rule->unused_tuple & BIT(INNER_DST_PORT) ?
+ 0 : cpu_to_be16(rule->tuples_mask.dst_port);
+
+ break;
+ case IPV6_USER_FLOW:
+ cpu_to_be32_array(fs->h_u.usr_ip6_spec.ip6src,
+ rule->tuples.src_ip, 4);
+ if (rule->unused_tuple & BIT(INNER_SRC_IP))
+ memset(fs->m_u.usr_ip6_spec.ip6src, 0, sizeof(int) * 4);
+ else
+ cpu_to_be32_array(fs->m_u.usr_ip6_spec.ip6src,
+ rule->tuples_mask.src_ip, 4);
+
+ cpu_to_be32_array(fs->h_u.usr_ip6_spec.ip6dst,
+ rule->tuples.dst_ip, 4);
+ if (rule->unused_tuple & BIT(INNER_DST_IP))
+ memset(fs->m_u.usr_ip6_spec.ip6dst, 0, sizeof(int) * 4);
+ else
+ cpu_to_be32_array(fs->m_u.usr_ip6_spec.ip6dst,
+ rule->tuples_mask.dst_ip, 4);
+
+ fs->h_u.usr_ip6_spec.l4_proto = rule->tuples.ip_proto;
+ fs->m_u.usr_ip6_spec.l4_proto =
+ rule->unused_tuple & BIT(INNER_IP_PROTO) ?
+ 0 : rule->tuples_mask.ip_proto;
+
+ break;
+ case ETHER_FLOW:
+ ether_addr_copy(fs->h_u.ether_spec.h_source,
+ rule->tuples.src_mac);
+ if (rule->unused_tuple & BIT(INNER_SRC_MAC))
+ eth_zero_addr(fs->m_u.ether_spec.h_source);
+ else
+ ether_addr_copy(fs->m_u.ether_spec.h_source,
+ rule->tuples_mask.src_mac);
+
+ ether_addr_copy(fs->h_u.ether_spec.h_dest,
+ rule->tuples.dst_mac);
+ if (rule->unused_tuple & BIT(INNER_DST_MAC))
+ eth_zero_addr(fs->m_u.ether_spec.h_dest);
+ else
+ ether_addr_copy(fs->m_u.ether_spec.h_dest,
+ rule->tuples_mask.dst_mac);
+
+ fs->h_u.ether_spec.h_proto =
+ cpu_to_be16(rule->tuples.ether_proto);
+ fs->m_u.ether_spec.h_proto =
+ rule->unused_tuple & BIT(INNER_ETH_TYPE) ?
+ 0 : cpu_to_be16(rule->tuples_mask.ether_proto);
+
+ break;
+ default:
+ return -EOPNOTSUPP;
+ }
+
+ if (fs->flow_type & FLOW_EXT) {
+ fs->h_ext.vlan_tci = cpu_to_be16(rule->tuples.vlan_tag1);
+ fs->m_ext.vlan_tci =
+ rule->unused_tuple & BIT(INNER_VLAN_TAG_FST) ?
+ cpu_to_be16(VLAN_VID_MASK) :
+ cpu_to_be16(rule->tuples_mask.vlan_tag1);
+ }
+
+ if (fs->flow_type & FLOW_MAC_EXT) {
+ ether_addr_copy(fs->h_ext.h_dest, rule->tuples.dst_mac);
+ if (rule->unused_tuple & BIT(INNER_DST_MAC))
+ eth_zero_addr(fs->m_u.ether_spec.h_dest);
+ else
+ ether_addr_copy(fs->m_u.ether_spec.h_dest,
+ rule->tuples_mask.dst_mac);
+ }
+
+ if (rule->action == HCLGE_FD_ACTION_DROP_PACKET) {
+ fs->ring_cookie = RX_CLS_FLOW_DISC;
+ } else {
+ u64 vf_id;
+
+ fs->ring_cookie = rule->queue_id;
+ vf_id = rule->vf_id;
+ vf_id <<= ETHTOOL_RX_FLOW_SPEC_RING_VF_OFF;
+ fs->ring_cookie |= vf_id;
+ }
+
+ return 0;
+}
+
+static int hclge_get_all_rules(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd, u32 *rule_locs)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+ struct hclge_fd_rule *rule;
+ struct hlist_node *node2;
+ int cnt = 0;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ cmd->data = hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1];
+
+ hlist_for_each_entry_safe(rule, node2,
+ &hdev->fd_rule_list, rule_node) {
+ if (cnt == cmd->rule_cnt)
+ return -EMSGSIZE;
+
+ rule_locs[cnt] = rule->location;
+ cnt++;
+ }
+
+ cmd->rule_cnt = cnt;
+
+ return 0;
+}
+
static void hclge_cfg_mac_mode(struct hclge_dev *hdev, bool enable)
{
struct hclge_desc desc;
@@ -7038,6 +7274,9 @@ static const struct hnae3_ae_ops hclge_ops = {
.get_link_mode = hclge_get_link_mode,
.add_fd_entry = hclge_add_fd_entry,
.del_fd_entry = hclge_del_fd_entry,
+ .get_fd_rule_cnt = hclge_get_fd_rule_cnt,
+ .get_fd_rule_info = hclge_get_fd_rule_info,
+ .get_fd_all_rules = hclge_get_all_rules,
};
static struct hnae3_ae_algo ae_algo = {
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 3/7] net: hns3: Add support for rule add/delete for flow director
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm, Jian Shen
In-Reply-To: <20181001114647.15504-1-salil.mehta@huawei.com>
From: Jian Shen <shenjian15@huawei.com>
This patch adds support for add and delete rule by ethtool commands.
HNS3 driver supports several flow types, include ETHER_FLOW,
IP_USER_FLOW, TCP_V4_FLOW, UDP_V4_FLOW, SCTP_V4_FLOW, IPV6_USER_FLOW,
TCP_V6_FLOW, UDP_V6_FLOW and SCTP_V6_FLOW.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 4 +
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 14 +-
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 501 +++++++++++++++++++++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 2 +
4 files changed, 519 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index d3e5190..262bb736 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -416,6 +416,10 @@ struct hnae3_ae_ops {
void (*get_link_mode)(struct hnae3_handle *handle,
unsigned long *supported,
unsigned long *advertising);
+ int (*add_fd_entry)(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd);
+ int (*del_fd_entry)(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd);
};
struct hnae3_dcb_ops {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 8ad2c3e..59cbf78 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -795,12 +795,22 @@ static int hns3_set_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd)
{
struct hnae3_handle *h = hns3_get_handle(netdev);
- if (!h->ae_algo || !h->ae_algo->ops || !h->ae_algo->ops->set_rss_tuple)
+ if (!h->ae_algo || !h->ae_algo->ops)
return -EOPNOTSUPP;
switch (cmd->cmd) {
case ETHTOOL_SRXFH:
- return h->ae_algo->ops->set_rss_tuple(h, cmd);
+ if (h->ae_algo->ops->set_rss_tuple)
+ return h->ae_algo->ops->set_rss_tuple(h, cmd);
+ return -EOPNOTSUPP;
+ case ETHTOOL_SRXCLSRLINS:
+ if (h->ae_algo->ops->add_fd_entry)
+ return h->ae_algo->ops->add_fd_entry(h, cmd);
+ return -EOPNOTSUPP;
+ case ETHTOOL_SRXCLSRLDEL:
+ if (h->ae_algo->ops->del_fd_entry)
+ return h->ae_algo->ops->del_fd_entry(h, cmd);
+ return -EOPNOTSUPP;
default:
return -EOPNOTSUPP;
}
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 030898d..43f5caa 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -3800,6 +3800,505 @@ static int hclge_config_action(struct hclge_dev *hdev, u8 stage,
return hclge_fd_ad_config(hdev, stage, ad_data.ad_id, &ad_data);
}
+static int hclge_fd_check_spec(struct hclge_dev *hdev,
+ struct ethtool_rx_flow_spec *fs, u32 *unused)
+{
+ struct ethtool_tcpip4_spec *tcp_ip4_spec;
+ struct ethtool_usrip4_spec *usr_ip4_spec;
+ struct ethtool_tcpip6_spec *tcp_ip6_spec;
+ struct ethtool_usrip6_spec *usr_ip6_spec;
+ struct ethhdr *ether_spec;
+
+ if (fs->location >= hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1])
+ return -EINVAL;
+
+ if (!(fs->flow_type & hdev->fd_cfg.proto_support))
+ return -EOPNOTSUPP;
+
+ if ((fs->flow_type & FLOW_EXT) &&
+ (fs->h_ext.data[0] != 0 || fs->h_ext.data[1] != 0)) {
+ dev_err(&hdev->pdev->dev, "user-def bytes are not supported\n");
+ return -EOPNOTSUPP;
+ }
+
+ switch (fs->flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) {
+ case SCTP_V4_FLOW:
+ case TCP_V4_FLOW:
+ case UDP_V4_FLOW:
+ tcp_ip4_spec = &fs->h_u.tcp_ip4_spec;
+ *unused |= BIT(INNER_SRC_MAC) | BIT(INNER_DST_MAC);
+
+ if (!tcp_ip4_spec->ip4src)
+ *unused |= BIT(INNER_SRC_IP);
+
+ if (!tcp_ip4_spec->ip4dst)
+ *unused |= BIT(INNER_DST_IP);
+
+ if (!tcp_ip4_spec->psrc)
+ *unused |= BIT(INNER_SRC_PORT);
+
+ if (!tcp_ip4_spec->pdst)
+ *unused |= BIT(INNER_DST_PORT);
+
+ if (!tcp_ip4_spec->tos)
+ *unused |= BIT(INNER_IP_TOS);
+
+ break;
+ case IP_USER_FLOW:
+ usr_ip4_spec = &fs->h_u.usr_ip4_spec;
+ *unused |= BIT(INNER_SRC_MAC) | BIT(INNER_DST_MAC) |
+ BIT(INNER_SRC_PORT) | BIT(INNER_DST_PORT);
+
+ if (!usr_ip4_spec->ip4src)
+ *unused |= BIT(INNER_SRC_IP);
+
+ if (!usr_ip4_spec->ip4dst)
+ *unused |= BIT(INNER_DST_IP);
+
+ if (!usr_ip4_spec->tos)
+ *unused |= BIT(INNER_IP_TOS);
+
+ if (!usr_ip4_spec->proto)
+ *unused |= BIT(INNER_IP_PROTO);
+
+ if (usr_ip4_spec->l4_4_bytes)
+ return -EOPNOTSUPP;
+
+ if (usr_ip4_spec->ip_ver != ETH_RX_NFC_IP4)
+ return -EOPNOTSUPP;
+
+ break;
+ case SCTP_V6_FLOW:
+ case TCP_V6_FLOW:
+ case UDP_V6_FLOW:
+ tcp_ip6_spec = &fs->h_u.tcp_ip6_spec;
+ *unused |= BIT(INNER_SRC_MAC) | BIT(INNER_DST_MAC) |
+ BIT(INNER_IP_TOS);
+
+ if (!tcp_ip6_spec->ip6src[0] && !tcp_ip6_spec->ip6src[1] &&
+ !tcp_ip6_spec->ip6src[2] && !tcp_ip6_spec->ip6src[3])
+ *unused |= BIT(INNER_SRC_IP);
+
+ if (!tcp_ip6_spec->ip6dst[0] && !tcp_ip6_spec->ip6dst[1] &&
+ !tcp_ip6_spec->ip6dst[2] && !tcp_ip6_spec->ip6dst[3])
+ *unused |= BIT(INNER_DST_IP);
+
+ if (!tcp_ip6_spec->psrc)
+ *unused |= BIT(INNER_SRC_PORT);
+
+ if (!tcp_ip6_spec->pdst)
+ *unused |= BIT(INNER_DST_PORT);
+
+ if (tcp_ip6_spec->tclass)
+ return -EOPNOTSUPP;
+
+ break;
+ case IPV6_USER_FLOW:
+ usr_ip6_spec = &fs->h_u.usr_ip6_spec;
+ *unused |= BIT(INNER_SRC_MAC) | BIT(INNER_DST_MAC) |
+ BIT(INNER_IP_TOS) | BIT(INNER_SRC_PORT) |
+ BIT(INNER_DST_PORT);
+
+ if (!usr_ip6_spec->ip6src[0] && !usr_ip6_spec->ip6src[1] &&
+ !usr_ip6_spec->ip6src[2] && !usr_ip6_spec->ip6src[3])
+ *unused |= BIT(INNER_SRC_IP);
+
+ if (!usr_ip6_spec->ip6dst[0] && !usr_ip6_spec->ip6dst[1] &&
+ !usr_ip6_spec->ip6dst[2] && !usr_ip6_spec->ip6dst[3])
+ *unused |= BIT(INNER_DST_IP);
+
+ if (!usr_ip6_spec->l4_proto)
+ *unused |= BIT(INNER_IP_PROTO);
+
+ if (usr_ip6_spec->tclass)
+ return -EOPNOTSUPP;
+
+ if (usr_ip6_spec->l4_4_bytes)
+ return -EOPNOTSUPP;
+
+ break;
+ case ETHER_FLOW:
+ ether_spec = &fs->h_u.ether_spec;
+ *unused |= BIT(INNER_SRC_IP) | BIT(INNER_DST_IP) |
+ BIT(INNER_SRC_PORT) | BIT(INNER_DST_PORT) |
+ BIT(INNER_IP_TOS) | BIT(INNER_IP_PROTO);
+
+ if (is_zero_ether_addr(ether_spec->h_source))
+ *unused |= BIT(INNER_SRC_MAC);
+
+ if (is_zero_ether_addr(ether_spec->h_dest))
+ *unused |= BIT(INNER_DST_MAC);
+
+ if (!ether_spec->h_proto)
+ *unused |= BIT(INNER_ETH_TYPE);
+
+ break;
+ default:
+ return -EOPNOTSUPP;
+ }
+
+ if ((fs->flow_type & FLOW_EXT)) {
+ if (fs->h_ext.vlan_etype)
+ return -EOPNOTSUPP;
+ if (!fs->h_ext.vlan_tci)
+ *unused |= BIT(INNER_VLAN_TAG_FST);
+
+ if (fs->m_ext.vlan_tci) {
+ if (be16_to_cpu(fs->h_ext.vlan_tci) >= VLAN_N_VID)
+ return -EINVAL;
+ }
+ } else {
+ *unused |= BIT(INNER_VLAN_TAG_FST);
+ }
+
+ if (fs->flow_type & FLOW_MAC_EXT) {
+ if (!(hdev->fd_cfg.proto_support & ETHER_FLOW))
+ return -EOPNOTSUPP;
+
+ if (is_zero_ether_addr(fs->h_ext.h_dest))
+ *unused |= BIT(INNER_DST_MAC);
+ else
+ *unused &= ~(BIT(INNER_DST_MAC));
+ }
+
+ return 0;
+}
+
+static bool hclge_fd_rule_exist(struct hclge_dev *hdev, u16 location)
+{
+ struct hclge_fd_rule *rule = NULL;
+ struct hlist_node *node2;
+
+ hlist_for_each_entry_safe(rule, node2, &hdev->fd_rule_list, rule_node) {
+ if (rule->location >= location)
+ break;
+ }
+
+ return rule && rule->location == location;
+}
+
+static int hclge_fd_update_rule_list(struct hclge_dev *hdev,
+ struct hclge_fd_rule *new_rule,
+ u16 location,
+ bool is_add)
+{
+ struct hclge_fd_rule *rule = NULL, *parent = NULL;
+ struct hlist_node *node2;
+
+ if (is_add && !new_rule)
+ return -EINVAL;
+
+ hlist_for_each_entry_safe(rule, node2,
+ &hdev->fd_rule_list, rule_node) {
+ if (rule->location >= location)
+ break;
+ parent = rule;
+ }
+
+ if (rule && rule->location == location) {
+ hlist_del(&rule->rule_node);
+ kfree(rule);
+ hdev->hclge_fd_rule_num--;
+
+ if (!is_add)
+ return 0;
+
+ } else if (!is_add) {
+ dev_err(&hdev->pdev->dev,
+ "delete fail, rule %d is inexistent\n",
+ location);
+ return -EINVAL;
+ }
+
+ INIT_HLIST_NODE(&new_rule->rule_node);
+
+ if (parent)
+ hlist_add_behind(&new_rule->rule_node, &parent->rule_node);
+ else
+ hlist_add_head(&new_rule->rule_node, &hdev->fd_rule_list);
+
+ hdev->hclge_fd_rule_num++;
+
+ return 0;
+}
+
+static int hclge_fd_get_tuple(struct hclge_dev *hdev,
+ struct ethtool_rx_flow_spec *fs,
+ struct hclge_fd_rule *rule)
+{
+ u32 flow_type = fs->flow_type & ~(FLOW_EXT | FLOW_MAC_EXT);
+
+ switch (flow_type) {
+ case SCTP_V4_FLOW:
+ case TCP_V4_FLOW:
+ case UDP_V4_FLOW:
+ rule->tuples.src_ip[3] =
+ be32_to_cpu(fs->h_u.tcp_ip4_spec.ip4src);
+ rule->tuples_mask.src_ip[3] =
+ be32_to_cpu(fs->m_u.tcp_ip4_spec.ip4src);
+
+ rule->tuples.dst_ip[3] =
+ be32_to_cpu(fs->h_u.tcp_ip4_spec.ip4dst);
+ rule->tuples_mask.dst_ip[3] =
+ be32_to_cpu(fs->m_u.tcp_ip4_spec.ip4dst);
+
+ rule->tuples.src_port = be16_to_cpu(fs->h_u.tcp_ip4_spec.psrc);
+ rule->tuples_mask.src_port =
+ be16_to_cpu(fs->m_u.tcp_ip4_spec.psrc);
+
+ rule->tuples.dst_port = be16_to_cpu(fs->h_u.tcp_ip4_spec.pdst);
+ rule->tuples_mask.dst_port =
+ be16_to_cpu(fs->m_u.tcp_ip4_spec.pdst);
+
+ rule->tuples.ip_tos = fs->h_u.tcp_ip4_spec.tos;
+ rule->tuples_mask.ip_tos = fs->m_u.tcp_ip4_spec.tos;
+
+ rule->tuples.ether_proto = ETH_P_IP;
+ rule->tuples_mask.ether_proto = 0xFFFF;
+
+ break;
+ case IP_USER_FLOW:
+ rule->tuples.src_ip[3] =
+ be32_to_cpu(fs->h_u.usr_ip4_spec.ip4src);
+ rule->tuples_mask.src_ip[3] =
+ be32_to_cpu(fs->m_u.usr_ip4_spec.ip4src);
+
+ rule->tuples.dst_ip[3] =
+ be32_to_cpu(fs->h_u.usr_ip4_spec.ip4dst);
+ rule->tuples_mask.dst_ip[3] =
+ be32_to_cpu(fs->m_u.usr_ip4_spec.ip4dst);
+
+ rule->tuples.ip_tos = fs->h_u.usr_ip4_spec.tos;
+ rule->tuples_mask.ip_tos = fs->m_u.usr_ip4_spec.tos;
+
+ rule->tuples.ip_proto = fs->h_u.usr_ip4_spec.proto;
+ rule->tuples_mask.ip_proto = fs->m_u.usr_ip4_spec.proto;
+
+ rule->tuples.ether_proto = ETH_P_IP;
+ rule->tuples_mask.ether_proto = 0xFFFF;
+
+ break;
+ case SCTP_V6_FLOW:
+ case TCP_V6_FLOW:
+ case UDP_V6_FLOW:
+ be32_to_cpu_array(rule->tuples.src_ip,
+ fs->h_u.tcp_ip6_spec.ip6src, 4);
+ be32_to_cpu_array(rule->tuples_mask.src_ip,
+ fs->m_u.tcp_ip6_spec.ip6src, 4);
+
+ be32_to_cpu_array(rule->tuples.dst_ip,
+ fs->h_u.tcp_ip6_spec.ip6dst, 4);
+ be32_to_cpu_array(rule->tuples_mask.dst_ip,
+ fs->m_u.tcp_ip6_spec.ip6dst, 4);
+
+ rule->tuples.src_port = be16_to_cpu(fs->h_u.tcp_ip6_spec.psrc);
+ rule->tuples_mask.src_port =
+ be16_to_cpu(fs->m_u.tcp_ip6_spec.psrc);
+
+ rule->tuples.dst_port = be16_to_cpu(fs->h_u.tcp_ip6_spec.pdst);
+ rule->tuples_mask.dst_port =
+ be16_to_cpu(fs->m_u.tcp_ip6_spec.pdst);
+
+ rule->tuples.ether_proto = ETH_P_IPV6;
+ rule->tuples_mask.ether_proto = 0xFFFF;
+
+ break;
+ case IPV6_USER_FLOW:
+ be32_to_cpu_array(rule->tuples.src_ip,
+ fs->h_u.usr_ip6_spec.ip6src, 4);
+ be32_to_cpu_array(rule->tuples_mask.src_ip,
+ fs->m_u.usr_ip6_spec.ip6src, 4);
+
+ be32_to_cpu_array(rule->tuples.dst_ip,
+ fs->h_u.usr_ip6_spec.ip6dst, 4);
+ be32_to_cpu_array(rule->tuples_mask.dst_ip,
+ fs->m_u.usr_ip6_spec.ip6dst, 4);
+
+ rule->tuples.ip_proto = fs->h_u.usr_ip6_spec.l4_proto;
+ rule->tuples_mask.ip_proto = fs->m_u.usr_ip6_spec.l4_proto;
+
+ rule->tuples.ether_proto = ETH_P_IPV6;
+ rule->tuples_mask.ether_proto = 0xFFFF;
+
+ break;
+ case ETHER_FLOW:
+ ether_addr_copy(rule->tuples.src_mac,
+ fs->h_u.ether_spec.h_source);
+ ether_addr_copy(rule->tuples_mask.src_mac,
+ fs->m_u.ether_spec.h_source);
+
+ ether_addr_copy(rule->tuples.dst_mac,
+ fs->h_u.ether_spec.h_dest);
+ ether_addr_copy(rule->tuples_mask.dst_mac,
+ fs->m_u.ether_spec.h_dest);
+
+ rule->tuples.ether_proto =
+ be16_to_cpu(fs->h_u.ether_spec.h_proto);
+ rule->tuples_mask.ether_proto =
+ be16_to_cpu(fs->m_u.ether_spec.h_proto);
+
+ break;
+ default:
+ return -EOPNOTSUPP;
+ }
+
+ switch (flow_type) {
+ case SCTP_V4_FLOW:
+ case SCTP_V6_FLOW:
+ rule->tuples.ip_proto = IPPROTO_SCTP;
+ rule->tuples_mask.ip_proto = 0xFF;
+ break;
+ case TCP_V4_FLOW:
+ case TCP_V6_FLOW:
+ rule->tuples.ip_proto = IPPROTO_TCP;
+ rule->tuples_mask.ip_proto = 0xFF;
+ break;
+ case UDP_V4_FLOW:
+ case UDP_V6_FLOW:
+ rule->tuples.ip_proto = IPPROTO_UDP;
+ rule->tuples_mask.ip_proto = 0xFF;
+ break;
+ default:
+ break;
+ }
+
+ if ((fs->flow_type & FLOW_EXT)) {
+ rule->tuples.vlan_tag1 = be16_to_cpu(fs->h_ext.vlan_tci);
+ rule->tuples_mask.vlan_tag1 = be16_to_cpu(fs->m_ext.vlan_tci);
+ }
+
+ if (fs->flow_type & FLOW_MAC_EXT) {
+ ether_addr_copy(rule->tuples.dst_mac, fs->h_ext.h_dest);
+ ether_addr_copy(rule->tuples_mask.dst_mac, fs->m_ext.h_dest);
+ }
+
+ return 0;
+}
+
+static int hclge_add_fd_entry(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+ u16 dst_vport_id = 0, q_index = 0;
+ struct ethtool_rx_flow_spec *fs;
+ struct hclge_fd_rule *rule;
+ u32 unused = 0;
+ u8 action;
+ int ret;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ if (!hdev->fd_cfg.fd_en) {
+ dev_warn(&hdev->pdev->dev,
+ "Please enable flow director first\n");
+ return -EOPNOTSUPP;
+ }
+
+ fs = (struct ethtool_rx_flow_spec *)&cmd->fs;
+
+ ret = hclge_fd_check_spec(hdev, fs, &unused);
+ if (ret) {
+ dev_err(&hdev->pdev->dev, "Check fd spec failed\n");
+ return ret;
+ }
+
+ if (fs->ring_cookie == RX_CLS_FLOW_DISC) {
+ action = HCLGE_FD_ACTION_DROP_PACKET;
+ } else {
+ u32 ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
+ u8 vf = ethtool_get_flow_spec_ring_vf(fs->ring_cookie);
+ u16 tqps;
+
+ dst_vport_id = vf ? hdev->vport[vf].vport_id : vport->vport_id;
+ tqps = vf ? hdev->vport[vf].alloc_tqps : vport->alloc_tqps;
+
+ if (ring >= tqps) {
+ dev_err(&hdev->pdev->dev,
+ "Error: queue id (%d) > max tqp num (%d)\n",
+ ring, tqps - 1);
+ return -EINVAL;
+ }
+
+ if (vf > hdev->num_req_vfs) {
+ dev_err(&hdev->pdev->dev,
+ "Error: vf id (%d) > max vf num (%d)\n",
+ vf, hdev->num_req_vfs);
+ return -EINVAL;
+ }
+
+ action = HCLGE_FD_ACTION_ACCEPT_PACKET;
+ q_index = ring;
+ }
+
+ rule = kzalloc(sizeof(*rule), GFP_KERNEL);
+ if (!rule)
+ return -ENOMEM;
+
+ ret = hclge_fd_get_tuple(hdev, fs, rule);
+ if (ret)
+ goto free_rule;
+
+ rule->flow_type = fs->flow_type;
+
+ rule->location = fs->location;
+ rule->unused_tuple = unused;
+ rule->vf_id = dst_vport_id;
+ rule->queue_id = q_index;
+ rule->action = action;
+
+ ret = hclge_config_action(hdev, HCLGE_FD_STAGE_1, rule);
+ if (ret)
+ goto free_rule;
+
+ ret = hclge_config_key(hdev, HCLGE_FD_STAGE_1, rule);
+ if (ret)
+ goto free_rule;
+
+ ret = hclge_fd_update_rule_list(hdev, rule, fs->location, true);
+ if (ret)
+ goto free_rule;
+
+ return ret;
+
+free_rule:
+ kfree(rule);
+ return ret;
+}
+
+static int hclge_del_fd_entry(struct hnae3_handle *handle,
+ struct ethtool_rxnfc *cmd)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+ struct ethtool_rx_flow_spec *fs;
+ int ret;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ fs = (struct ethtool_rx_flow_spec *)&cmd->fs;
+
+ if (fs->location >= hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1])
+ return -EINVAL;
+
+ if (!hclge_fd_rule_exist(hdev, fs->location)) {
+ dev_err(&hdev->pdev->dev,
+ "Delete fail, rule %d is inexistent\n",
+ fs->location);
+ return -ENOENT;
+ }
+
+ ret = hclge_fd_tcam_config(hdev, HCLGE_FD_STAGE_1, true,
+ fs->location, NULL, false);
+ if (ret)
+ return ret;
+
+ return hclge_fd_update_rule_list(hdev, NULL, fs->location,
+ false);
+}
+
static void hclge_cfg_mac_mode(struct hclge_dev *hdev, bool enable)
{
struct hclge_desc desc;
@@ -6537,6 +7036,8 @@ static const struct hnae3_ae_ops hclge_ops = {
.get_regs = hclge_get_regs,
.set_led_id = hclge_set_led_id,
.get_link_mode = hclge_get_link_mode,
+ .add_fd_entry = hclge_add_fd_entry,
+ .del_fd_entry = hclge_del_fd_entry,
};
static struct hnae3_ae_algo ae_algo = {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index 7f5a242..06adbdd 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -678,6 +678,8 @@ struct hclge_dev {
unsigned long vlan_table[VLAN_N_VID][BITS_TO_LONGS(HCLGE_VPORT_NUM)];
struct hclge_fd_cfg fd_cfg;
+ struct hlist_head fd_rule_list;
+ u16 hclge_fd_rule_num;
};
/* VPort level vlan tag configuration for TX direction */
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 5/7] net: hns3: Add reset handle for flow director
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm, Jian Shen
In-Reply-To: <20181001114647.15504-1-salil.mehta@huawei.com>
From: Jian Shen <shenjian15@huawei.com>
When doing reset, remove all entries in TCAM block, and keep flow
director rules list. After finishing reset, restore all entries.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 4 ++
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 30 ++++++++++
drivers/net/ethernet/hisilicon/hns3/hns3_enet.h | 5 ++
.../ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 64 ++++++++++++++++++++++
4 files changed, 103 insertions(+)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index fac84d8..e185d04 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -179,6 +179,7 @@ struct hnae3_ae_dev {
struct list_head node;
u32 flag;
enum hnae3_dev_type dev_type;
+ enum hnae3_reset_type reset_type;
void *priv;
};
@@ -420,12 +421,15 @@ struct hnae3_ae_ops {
struct ethtool_rxnfc *cmd);
int (*del_fd_entry)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd);
+ void (*del_all_fd_entries)(struct hnae3_handle *handle,
+ bool clear_list);
int (*get_fd_rule_cnt)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd);
int (*get_fd_rule_info)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd);
int (*get_fd_all_rules)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd, u32 *rule_locs);
+ int (*restore_fd_rules)(struct hnae3_handle *handle);
};
struct hnae3_dcb_ops {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 605f029..df4410e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1654,6 +1654,7 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
ae_dev->pdev = pdev;
ae_dev->flag = ent->driver_data;
ae_dev->dev_type = HNAE3_DEV_KNIC;
+ ae_dev->reset_type = HNAE3_NONE_RESET;
hns3_get_dev_capability(pdev, ae_dev);
pci_set_drvdata(pdev, ae_dev);
@@ -3150,6 +3151,25 @@ static void hns3_uninit_mac_addr(struct net_device *netdev)
h->ae_algo->ops->rm_uc_addr(h, netdev->dev_addr);
}
+static int hns3_restore_fd_rules(struct net_device *netdev)
+{
+ struct hnae3_handle *h = hns3_get_handle(netdev);
+ int ret = 0;
+
+ if (h->ae_algo->ops->restore_fd_rules)
+ ret = h->ae_algo->ops->restore_fd_rules(h);
+
+ return ret;
+}
+
+static void hns3_del_all_fd_rules(struct net_device *netdev, bool clear_list)
+{
+ struct hnae3_handle *h = hns3_get_handle(netdev);
+
+ if (h->ae_algo->ops->del_all_fd_entries)
+ h->ae_algo->ops->del_all_fd_entries(h, clear_list);
+}
+
static void hns3_nic_set_priv_ops(struct net_device *netdev)
{
struct hns3_nic_priv *priv = netdev_priv(netdev);
@@ -3561,6 +3581,8 @@ static int hns3_reset_notify_init_enet(struct hnae3_handle *handle)
if (!(handle->flags & HNAE3_SUPPORT_VF))
hns3_restore_vlan(netdev);
+ hns3_restore_fd_rules(netdev);
+
/* Carrier off reporting is important to ethtool even BEFORE open */
netif_carrier_off(netdev);
@@ -3581,6 +3603,7 @@ static int hns3_reset_notify_init_enet(struct hnae3_handle *handle)
static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
{
+ struct hnae3_ae_dev *ae_dev = pci_get_drvdata(handle->pdev);
struct net_device *netdev = handle->kinfo.netdev;
struct hns3_nic_priv *priv = netdev_priv(netdev);
int ret;
@@ -3601,6 +3624,13 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
hns3_uninit_mac_addr(netdev);
+ /* it is cumbersome for hardware to pick-and-choose rules for deletion
+ * from TCAM. Hence, for function reset software intervention is
+ * required to delete the rules
+ */
+ if (hns3_dev_ongoing_func_reset(ae_dev))
+ hns3_del_all_fd_rules(netdev, false);
+
return ret;
}
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
index 27d704f..ac881e8 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
@@ -585,6 +585,11 @@ static inline void hns3_write_reg(void __iomem *base, u32 reg, u32 value)
writel(value, reg_addr + reg);
}
+static inline bool hns3_dev_ongoing_func_reset(struct hnae3_ae_dev *ae_dev)
+{
+ return (ae_dev && (ae_dev->reset_type == HNAE3_FUNC_RESET));
+}
+
#define hns3_write_dev(a, reg, value) \
hns3_write_reg((a)->io_base, (reg), (value))
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 876c7ca..00ec4db 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -2513,8 +2513,13 @@ static void hclge_clear_reset_cause(struct hclge_dev *hdev)
static void hclge_reset(struct hclge_dev *hdev)
{
+ struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev);
struct hnae3_handle *handle;
+ /* Initialize ae_dev reset status as well, in case enet layer wants to
+ * know if device is undergoing reset
+ */
+ ae_dev->reset_type = hdev->reset_type;
/* perform reset of the stack & ae device for a client */
handle = &hdev->vport[0].nic;
rtnl_lock();
@@ -2535,6 +2540,7 @@ static void hclge_reset(struct hclge_dev *hdev)
hclge_notify_client(hdev, HNAE3_UP_CLIENT);
handle->last_reset_time = jiffies;
rtnl_unlock();
+ ae_dev->reset_type = HNAE3_NONE_RESET;
}
static void hclge_reset_event(struct hnae3_handle *handle)
@@ -4299,6 +4305,62 @@ static int hclge_del_fd_entry(struct hnae3_handle *handle,
false);
}
+static void hclge_del_all_fd_entries(struct hnae3_handle *handle,
+ bool clear_list)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+ struct hclge_fd_rule *rule;
+ struct hlist_node *node;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return;
+
+ if (clear_list) {
+ hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list,
+ rule_node) {
+ hclge_fd_tcam_config(hdev, HCLGE_FD_STAGE_1, true,
+ rule->location, NULL, false);
+ hlist_del(&rule->rule_node);
+ kfree(rule);
+ hdev->hclge_fd_rule_num--;
+ }
+ } else {
+ hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list,
+ rule_node)
+ hclge_fd_tcam_config(hdev, HCLGE_FD_STAGE_1, true,
+ rule->location, NULL, false);
+ }
+}
+
+static int hclge_restore_fd_entries(struct hnae3_handle *handle)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+ struct hclge_fd_rule *rule;
+ struct hlist_node *node;
+ int ret;
+
+ if (!hnae3_dev_fd_supported(hdev))
+ return -EOPNOTSUPP;
+
+ hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list, rule_node) {
+ ret = hclge_config_action(hdev, HCLGE_FD_STAGE_1, rule);
+ if (!ret)
+ ret = hclge_config_key(hdev, HCLGE_FD_STAGE_1, rule);
+
+ if (ret) {
+ dev_warn(&hdev->pdev->dev,
+ "Restore rule %d failed, remove it\n",
+ rule->location);
+ hlist_del(&rule->rule_node);
+ kfree(rule);
+ hdev->hclge_fd_rule_num--;
+ }
+ }
+ return 0;
+}
+
static int hclge_get_fd_rule_cnt(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd)
{
@@ -7274,9 +7336,11 @@ static const struct hnae3_ae_ops hclge_ops = {
.get_link_mode = hclge_get_link_mode,
.add_fd_entry = hclge_add_fd_entry,
.del_fd_entry = hclge_del_fd_entry,
+ .del_all_fd_entries = hclge_del_all_fd_entries,
.get_fd_rule_cnt = hclge_get_fd_rule_cnt,
.get_fd_rule_info = hclge_get_fd_rule_info,
.get_fd_all_rules = hclge_get_all_rules,
+ .restore_fd_rules = hclge_restore_fd_entries,
};
static struct hnae3_ae_algo ae_algo = {
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 6/7] net: hns3: Remove all flow director rules when unload hns3 driver
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm, Jian Shen
In-Reply-To: <20181001114647.15504-1-salil.mehta@huawei.com>
From: Jian Shen <shenjian15@huawei.com>
This patch removes all flow director rules when unload hns3 driver.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index df4410e..96f07ca 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -3286,6 +3286,8 @@ static void hns3_client_uninit(struct hnae3_handle *handle, bool reset)
if (netdev->reg_state != NETREG_UNINITIALIZED)
unregister_netdev(netdev);
+ hns3_del_all_fd_rules(netdev, true);
+
hns3_force_clear_all_rx_ring(handle);
ret = hns3_nic_uninit_vector_data(priv);
--
2.7.4
^ permalink raw reply related
* [PATCH net-next 7/7] net: hns3: Add support for enable/disable flow director
From: Salil Mehta @ 2018-10-01 11:46 UTC (permalink / raw)
To: davem
Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
linux-kernel, linuxarm, Jian Shen
In-Reply-To: <20181001114647.15504-1-salil.mehta@huawei.com>
From: Jian Shen <shenjian15@huawei.com>
This patch adds switch for flow director with ethtool command
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
drivers/net/ethernet/hisilicon/hns3/hnae3.h | 1 +
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 15 ++++++++++++++-
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 13 +++++++++++++
3 files changed, 28 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index e185d04..591ee2e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -430,6 +430,7 @@ struct hnae3_ae_ops {
int (*get_fd_all_rules)(struct hnae3_handle *handle,
struct ethtool_rxnfc *cmd, u32 *rule_locs);
int (*restore_fd_rules)(struct hnae3_handle *handle);
+ void (*enable_fd)(struct hnae3_handle *handle, bool enable);
};
struct hnae3_dcb_ops {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 96f07ca..c269256 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1285,6 +1285,13 @@ static int hns3_nic_set_features(struct net_device *netdev,
return ret;
}
+ if ((changed & NETIF_F_NTUPLE) && h->ae_algo->ops->enable_fd) {
+ if (features & NETIF_F_NTUPLE)
+ h->ae_algo->ops->enable_fd(h, true);
+ else
+ h->ae_algo->ops->enable_fd(h, false);
+ }
+
netdev->features = features;
return 0;
}
@@ -1770,8 +1777,14 @@ static void hns3_set_default_feature(struct net_device *netdev)
NETIF_F_GSO_GRE_CSUM | NETIF_F_GSO_UDP_TUNNEL |
NETIF_F_GSO_UDP_TUNNEL_CSUM | NETIF_F_SCTP_CRC;
- if (pdev->revision != 0x20)
+ if (pdev->revision >= 0x21) {
netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_FILTER;
+
+ if (!(h->flags & HNAE3_SUPPORT_VF)) {
+ netdev->hw_features |= NETIF_F_NTUPLE;
+ netdev->features |= NETIF_F_NTUPLE;
+ }
+ }
}
static int hns3_alloc_buffer(struct hns3_enet_ring *ring,
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 00ec4db..0ce7ccb 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -4597,6 +4597,18 @@ static int hclge_get_all_rules(struct hnae3_handle *handle,
return 0;
}
+static void hclge_enable_fd(struct hnae3_handle *handle, bool enable)
+{
+ struct hclge_vport *vport = hclge_get_vport(handle);
+ struct hclge_dev *hdev = vport->back;
+
+ hdev->fd_cfg.fd_en = enable;
+ if (!enable)
+ hclge_del_all_fd_entries(handle, false);
+ else
+ hclge_restore_fd_entries(handle);
+}
+
static void hclge_cfg_mac_mode(struct hclge_dev *hdev, bool enable)
{
struct hclge_desc desc;
@@ -7341,6 +7353,7 @@ static const struct hnae3_ae_ops hclge_ops = {
.get_fd_rule_info = hclge_get_fd_rule_info,
.get_fd_all_rules = hclge_get_all_rules,
.restore_fd_rules = hclge_restore_fd_entries,
+ .enable_fd = hclge_enable_fd,
};
static struct hnae3_ae_algo ae_algo = {
--
2.7.4
^ permalink raw reply related
* Re: [PATCH v4 bpf-next 00/10] bpf: per-cpu cgroup local storage
From: Daniel Borkmann @ 2018-10-01 19:08 UTC (permalink / raw)
To: Roman Gushchin, netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, Kernel Team
In-Reply-To: <20180928144452.5284-1-guro@fb.com>
On 09/28/2018 04:45 PM, Roman Gushchin wrote:
> This patchset implements per-cpu cgroup local storage and provides
> an example how per-cpu and shared cgroup local storage can be used
> for efficient accounting of network traffic.
>
> v4->v3:
> 1) incorporated Alexei's feedback
>
> v3->v2:
> 1) incorporated Song's feedback
> 2) rebased on top of current bpf-next
>
> v2->v1:
> 1) added a selftest implementing network counters
> 2) added a missing free() in cgroup local storage selftest
Applied to bpf-next, thanks!
^ permalink raw reply
* Re: WARNING in pcpu_alloc
From: Roman Gushchin @ 2018-10-01 19:21 UTC (permalink / raw)
To: Alexei Starovoitov
Cc: syzbot+18628320d3b14a5c459c@syzkaller.appspotmail.com,
Alexei Starovoitov, Daniel Borkmann, LKML, Network Development,
syzkaller-bugs
In-Reply-To: <CAADnVQ+DyZ5ze4gjfgQz+OATRE_Jd1b5D1osq8+8O3BL9vHzhg@mail.gmail.com>
Sure, will do soon. Thanks!
Sent from my iPhone
> On Oct 1, 2018, at 20:24, Alexei Starovoitov <alexei.starovoitov@gmail.com> wrote:
>
> On Mon, Oct 1, 2018 at 6:14 PM syzbot
> <syzbot+18628320d3b14a5c459c@syzkaller.appspotmail.com> wrote:
>>
>> Hello,
>>
>> syzbot found the following crash on:
>>
>> HEAD commit: c5a9f9817ba5 Merge branch 'bpf-per-cpu-cgroup-local-storage'
>> git tree: bpf-next
>> console output: https://urldefense.proofpoint.com/v2/url?u=https-3A__syzkaller.appspot.com_x_log.txt-3Fx-3D175f98e6400000&d=DwIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=RDq5scOAUEmnSy8tXTh8WldgNIIBkpJow65eBMO1X78&s=1gv4CbD91yt2XHXTnOVUyGN15DqWWFz3xYlo9ioGZEw&e=
>> kernel config: https://urldefense.proofpoint.com/v2/url?u=https-3A__syzkaller.appspot.com_x_.config-3Fx-3D443816db871edd66&d=DwIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=RDq5scOAUEmnSy8tXTh8WldgNIIBkpJow65eBMO1X78&s=EdREqNHdZup-52h5rZo0Mq59K3DYw1YVizBexkGikEk&e=
>> dashboard link: https://urldefense.proofpoint.com/v2/url?u=https-3A__syzkaller.appspot.com_bug-3Fextid-3D18628320d3b14a5c459c&d=DwIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=RDq5scOAUEmnSy8tXTh8WldgNIIBkpJow65eBMO1X78&s=QlmeF_tfsUIhzLLRKSwcR5ruQyqZA5wqow5kw_BRA6o&e=
>> compiler: gcc (GCC) 8.0.1 20180413 (experimental)
>> syz repro: https://urldefense.proofpoint.com/v2/url?u=https-3A__syzkaller.appspot.com_x_repro.syz-3Fx-3D129cbd1a400000&d=DwIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=RDq5scOAUEmnSy8tXTh8WldgNIIBkpJow65eBMO1X78&s=VDTRxfcVTxFp9LiiLsUYKevDzawNsYjRuqMHAZ7an2c&e=
>> C reproducer: https://urldefense.proofpoint.com/v2/url?u=https-3A__syzkaller.appspot.com_x_repro.c-3Fx-3D16685ac9400000&d=DwIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=RDq5scOAUEmnSy8tXTh8WldgNIIBkpJow65eBMO1X78&s=PIFdU3kSa5SEp7k0bav9JxpLGyBiPz84uRE25vyl5_0&e=
>>
>> IMPORTANT: if you fix the bug, please add the following tag to the commit:
>> Reported-by: syzbot+18628320d3b14a5c459c@syzkaller.appspotmail.com
>>
>> ------------[ cut here ]------------
>> illegal size (0) or align (8) for percpu allocation
>
> Roman,
>
> looks like
> if (attr->value_size > PAGE_SIZE)
> return ERR_PTR(-E2BIG);
> is not enough, since value_size==0 needs to be rejected as well.
> Could you prepare a patch?
^ permalink raw reply
* Re: [PATCH v3 net-next 07/12] net: ethernet: Add helper to remove a supported link mode
From: Simon Horman @ 2018-10-01 12:43 UTC (permalink / raw)
To: David Miller
Cc: andrew, netdev, f.fainelli, sergei.shtylyov, linux-renesas-soc
In-Reply-To: <20180926.200846.2016469227623990606.davem@davemloft.net>
On Wed, Sep 26, 2018 at 08:08:46PM -0700, David Miller wrote:
> From: Andrew Lunn <andrew@lunn.ch>
> Date: Mon, 24 Sep 2018 17:50:23 +0200
>
> > I submitted it to netdev in the usual way. I hope DaveM will accept
> > and merge it.
>
> Andrew did I miss your patch somehow?
>
> If so, sorry, please resend.
Hi Dave,
I believe you have the patch in net-next as:
65c5877f6462 ("ravb: Disable Pause Advertisement")
^ permalink raw reply
* Re: [PATCH net-next] rtnetlink: fix rtnl_fdb_dump() for shorter family headers
From: Mauricio Faria de Oliveira @ 2018-10-01 12:44 UTC (permalink / raw)
To: dsahern; +Cc: netdev, davem
In-Reply-To: <1cf63710-8ddb-2e21-4df4-f8687433f3fb@gmail.com>
On Sun, Sep 30, 2018 at 10:06 PM David Ahern <dsahern@gmail.com> wrote:
>
> On 9/28/18 1:35 PM, Mauricio Faria de Oliveira wrote:
> > Currently, rtnl_fdb_dump() assumes the family header is 'struct ifinfomsg',
> > which is not always true. For example, 'struct ndmsg' is used by iproute2
> > as well (in the 'ip neigh' command).
> >
> > The problem is, the function bails out early if nlmsg_parse() fails, which
> > does occur for iproute2 usage of 'struct ndmsg' because the payload length
> > is shorter than the family header alone (as 'struct ifinfomsg' is assumed).
> >
> > This breaks backward compatibility with userspace (different response) and
> > is a regression due to commit 0ff50e83b512 ("net: rtnetlink: bail out from
> > rtnl_fdb_dump() on parse error").
> ...
>
> >
> > Fixes: 0ff50e83b512 ("net: rtnetlink: bail out from rtnl_fdb_dump() on parse error")
> > Fixes: 5e6d24358799 ("bridge: netlink dump interface at par with brctl")
> > Reported-by: Aidan Obley <aobley@pivotal.io>
> > Signed-off-by: Mauricio Faria de Oliveira <mfo@canonical.com>
> > ---
> > P.S.: this may be 'net', but labeling as 'net-next' for possible relation to recent thread
> > [PATCH RFC net-next 0/5] rtnetlink: Add support for rigid checking of data in dump request
> >
> > net/core/rtnetlink.c | 15 ++++++++-------
> > 1 file changed, 8 insertions(+), 7 deletions(-)
> >
> > diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> > index 60c928894a78..9695a27cc9b9 100644
> > --- a/net/core/rtnetlink.c
> > +++ b/net/core/rtnetlink.c
> > @@ -3744,16 +3744,17 @@ static int rtnl_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb)
> > int err = 0;
> > int fidx = 0;
> >
> > - err = nlmsg_parse(cb->nlh, sizeof(struct ifinfomsg), tb,
> > - IFLA_MAX, ifla_policy, NULL);
> > - if (err < 0) {
> > - return -EINVAL;
> > - } else if (err == 0) {
> > + /* The family header may _not_ be struct ifinfomsg
> > + * (e.g., struct ndmsg). Usage of the ifm pointer
> > + * must check payload length (e.g., nlmsg_parse()).
> > + */
> > + if (nlmsg_parse(cb->nlh, sizeof(struct ifinfomsg), tb,
> > + IFLA_MAX, ifla_policy, NULL) == 0) {
> > if (tb[IFLA_MASTER])
> > br_idx = nla_get_u32(tb[IFLA_MASTER]);
> > - }
> >
> > - brport_idx = ifm->ifi_index;
> > + brport_idx = ifm->ifi_index;
> > + }
> >
> > if (br_idx) {
> > br_dev = __dev_get_by_index(net, br_idx);
> >
>
David, thanks for reviewing this.
> I suspect rtnl_fdb_dump is forever stuck with the ifinfomsg struct as
> the header if any kernel side filtering is to be done. [snip]
Why exactly? I understand currently there may be little information
to distinguish family headers, but if it comes down to certain attributes
the function expects/uses, that can be checked if nlmsg_parse() is OK.
Otherwise, if it comes down to some struct field that is not common
between both structs, then.. well. Maybe something else/new.
> [...] As for the change
> above, I suggest something like this:
>
> /* if header struct is ndmsg, no attributes can be appended */
> if (nlmsg_len(nlh) != sizeof(struct ndmsg)) {
> current ifinfomsg based code
> }
>
Hm, not sure -- ndmsg can be used with attributes too in iproute2, e.g., 'dev'.
In that case, the payload size is different/greater, and even makes
the length check pass:
$ ip --family bridge neigh
[ 585.111034] DEBUG: rtnl_fdb_dump():3749 nlmsg_len(nlh) 12
RTNETLINK answers: Invalid argument
Dump terminated
$ ip --family bridge neigh show dev ens3
[ 540.231443] DEBUG: rtnl_fdb_dump():3749 nlmsg_len(nlh) 20
[ 540.234433] DEBUG: rtnl_fdb_dump():3749 nlmsg_len(nlh) 20
lladdr 33:33:00:00:00:01 PERMANENT
lladdr 01:00:5e:00:00:01 PERMANENT
lladdr 33:33:ff:e9:9d:60 PERMANENT
> We certainly do not want to ignore parse failures.
Well, if there are no attributes, that shouldn't be a problem,
but unfortunately that's what happens in this case :- (
Now, given the example above makes the attribute parsing pass,
that allows the payload interpretation as ifm struct to be used.
Fortunately this field is commonly aligned on both structs,
and is not used in ndmsg per iproute2 ipneigh.c do_show_or_flush().
But this might pose problems in the future if things change.
If you could explain your thoughts a bit more about it, it would be
really great.
Thanks again,
--
Mauricio Faria de Oliveira
^ permalink raw reply
* Re: [PATCH v5 13/15] octeontx2-af: Add support for CGX link management
From: Arnd Bergmann @ 2018-10-01 12:46 UTC (permalink / raw)
To: Sunil Kovvuri
Cc: Networking, David Miller, linux-soc, lcherian, nmani, sgoutham
In-Reply-To: <1538393791-26251-14-git-send-email-sunil.kovvuri@gmail.com>
On Mon, Oct 1, 2018 at 1:37 PM <sunil.kovvuri@gmail.com> wrote:
> + * Not applicable for commands :
> + * CGX_CMD_LINK_BRING_UP/DOWN/CGX_EVT_LINK_CHANGE
> + * check struct cgx_lnk_sts comments
> + */
> +struct cgx_err_sts_s {
> +#if defined(__BIG_ENDIAN_BITFIELD)
> + uint64_t reserved2:35;
> + uint64_t type:10;
> + uint64_t reserved1:9;
> +#else
> + uint64_t reserved1:9;
> + uint64_t type:10; /* cgx_error_type */
> + uint64_t reserved2:35;
> +#endif
> +};
> +
> +/* Response to cmd ID as CGX_CMD_GET_FW_VER with cmd status as
> + * CGX_STAT_SUCCESS
> + */
> +struct cgx_ver_s {
> +#if defined(__BIG_ENDIAN_BITFIELD)
> + uint64_t reserved2:47;
> + uint64_t minor_ver:4;
> + uint64_t major_ver:4;
> + uint64_t reserved1:9;
> +#else
> + uint64_t reserved1:9;
> + uint64_t major_ver:4;
> + uint64_t minor_ver:4;
> + uint64_t reserved2:47;
> +#endif
> +};
This is not exactly what I meant in my comment. I think this may work
correctly on all architectures that we support in Linux, but my point
was that the layout of bit fields within a structure is completely up
to the implementation, so it's really better to just avoid it entirely.
Note that the data is being read with a call like
ereg.val = cgx_read(cgx, lmac->lmac_id, CGX_EVENT_REG);
which in turn is implemented using readq(), and that already
performs a byte swap in order to do the right thing on big-endian
architectures. The non-reserved bitfields in the two structures
above span byte boundaries, so I can't even wrap my head around
where the respective bits are supposed to be for all combinations
of __BIG_ENDIAN/__LITTLE_ENDIAN and __BIG_ENDIAN_BITFIELD/
__LITTLE_ENDIAN_BITFIELD.
Arnd
^ permalink raw reply
* Re: [net-next, PATCH 0/2, v3] net: socionext: XDP support
From: Björn Töpel @ 2018-10-01 12:48 UTC (permalink / raw)
To: Ilias Apalodimas, netdev, jaswinder.singh
Cc: ard.biesheuvel, masami.hiramatsu, arnd, magnus.karlsson, brouer,
daniel, ast, jesus.sanchez-palencia, vinicius.gomes,
makita.toshiaki
In-Reply-To: <1538220482-16129-1-git-send-email-ilias.apalodimas@linaro.org>
On 2018-09-29 13:28, Ilias Apalodimas wrote:
> This patch series adds AF_XDP support socionext netsec driver
This series adds *XDP* support and as a result, the AF_XDP batteries
are included. ;-)
Björn
> In addition the new dma allocation scheme offers a 10% boost on Rx
> pps rate using 64b packets
>
> - patch [1/2]: Use a different allocation scheme for Rx DMA buffers to
> prepare the driver for AF_XDP support
> - patch [2/2]: Add XDP support without zero-copy
>
> test and performance numbers (64b packets):
> -------------------------------------------
> - Normal SKBs on Rx: ~217kpps
> test: pktgen -> intel i210 -> netsec -> XDP_TX/XDP_REDIRECT
> - XDP_TX: 320kpps
> - XDP_REDIRECT: 320kpps
>
> qemu -> pktgen -> virtio -> ndo_xdp_xmit -> netsec
> - ndo_xdp_xmit: Could not send more than 120kpps. Interface forwarded that
> with success
>
> Changes since v2:
> - Always allocate Rx buffers with XDP_PACKET_HEADROOM
>
> Björn Töpel:
> - Added locking in the Tx queue
>
> Jesper Dangaard Brouer:
> - Added support for .ndo_xdp_xmit
> - XDP_TX does not flush every packet
>
> Changes since v1:
> - patch [2/2]:
> Toshiaki Makita:
> - Added XDP_PACKET_HEADROOM
> - Fixed a bug on XDP_PASS case
> - Using extact for error messaging instead of netdev_warn, when
> trying to setup XDP
>
> Ilias Apalodimas (2):
> net: socionext: different approach on DMA
> net: socionext: add XDP support
>
> drivers/net/ethernet/socionext/netsec.c | 541 +++++++++++++++++++++++++-------
> 1 file changed, 426 insertions(+), 115 deletions(-)
>
^ permalink raw reply
* [PATCH bpf-next v2 1/5] net: add umem reference in netdev{_rx}_queue
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
In-Reply-To: <1538398297-14862-1-git-send-email-magnus.karlsson@intel.com>
These references to the umem will be used to store information
on what kind of AF_XDP umem that is bound to a queue id, if any.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
include/linux/netdevice.h | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 1cbbf77..8318f79 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -609,6 +609,9 @@ struct netdev_queue {
/* Subordinate device that the queue has been assigned to */
struct net_device *sb_dev;
+#ifdef CONFIG_XDP_SOCKETS
+ struct xdp_umem *umem;
+#endif
/*
* write-mostly part
*/
@@ -738,6 +741,9 @@ struct netdev_rx_queue {
struct kobject kobj;
struct net_device *dev;
struct xdp_rxq_info xdp_rxq;
+#ifdef CONFIG_XDP_SOCKETS
+ struct xdp_umem *umem;
+#endif
} ____cacheline_aligned_in_smp;
/*
--
2.7.4
^ permalink raw reply related
* [PATCH bpf-next v2 0/5] xsk: fix bug when trying to use both copy and zero-copy mode
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
Previously, the xsk code did not record which umem was bound to a
specific queue id. This was not required if all drivers were zero-copy
enabled as this had to be recorded in the driver anyway. So if a user
tried to bind two umems to the same queue, the driver would say
no. But if copy-mode was first enabled and then zero-copy mode (or the
reverse order), we mistakenly enabled both of them on the same umem
leading to buggy behavior. The main culprit for this is that we did
not store the association of umem to queue id in the copy case and
only relied on the driver reporting this. As this relation was not
stored in the driver for copy mode (it does not rely on the AF_XDP
NDOs), this obviously could not work.
This patch fixes the problem by always recording the umem to queue id
relationship in the netdev_queue and netdev_rx_queue structs. This way
we always know what kind of umem has been bound to a queue id and can
act appropriately at bind time. To make the bind semantics consistent
with ethtool queue manipulations and to facilitate the implementation
of drivers, we also forbid decreasing the number of queues/channels
with ethtool if there is an active AF_XDP socket in the set of queues
that are disabled.
Jakub, please take a look at your patches. The last one I had to
change slightly to make it fit with the new interface
xdp_get_umem_from_qid(). An added bonus with this function is that we,
in the future, can also use it from the driver to get a umem, thus
simplifying driver implementations (and later remove the umem from the
NDO completely). Björn will mail patches, at a later point in time,
using this in the i40e and ixgbe drivers, that removes a good chunk of
code from the ZC implementations. I also made your code aware of Tx
queues. If we create a socket that only has a Tx queue, then the queue
id will refer to a Tx queue id only and could be larger than the
available amount of Rx queues. Please take a look at it.
Differences against v1:
* Included patches from Jakub that forbids decreasing the number of active
queues if a queue to be deactivated has an AF_XDP socket. These have
been adapted somewhat to the new interfaces in patch 2.
* Removed redundant check against real_num_[rt]x_queue in xsk_bind
* Only need to test against real_num_[rt]x_queues in
xdp_clear_umem_at_qid.
Patch 1: Introduces a umem reference in the netdev_rx_queue and
netdev_queue structs.
Patch 2: Records which queue_id is bound to which umem and make sure
that you cannot bind two different umems to the same queue_id.
Patch 3: Pre patch to ethtool_set_channels.
Patch 4: Forbid decreasing the number of active queues if a deactivated
queue has an AF_XDP socket.
Patch 5: Simplify xdp_clear_umem_at_qid now when ethtool cannot deactivate
the queue id we are running on.
I based this patch set on bpf-next commit 5bf7a60b8e70 ("bpf: permit
CGROUP_DEVICE programs accessing helper bpf_get_current_cgroup_id()")
Thanks: Magnus
Jakub Kicinski (2):
ethtool: rename local variable max -> curr
ethtool: don't allow disabling queues with umem installed
Magnus Karlsson (3):
net: add umem reference in netdev{_rx}_queue
xsk: fix bug when trying to use both copy and zero-copy on one queue
id
xsk: simplify xdp_clear_umem_at_qid implementation
include/linux/netdevice.h | 6 ++++
include/net/xdp_sock.h | 7 ++++
net/core/ethtool.c | 23 +++++++++----
net/xdp/xdp_umem.c | 87 ++++++++++++++++++++++++++++++++---------------
net/xdp/xdp_umem.h | 2 +-
net/xdp/xsk.c | 7 ----
6 files changed, 91 insertions(+), 41 deletions(-)
^ permalink raw reply
* [PATCH bpf-next v2 2/5] xsk: fix bug when trying to use both copy and zero-copy on one queue id
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
In-Reply-To: <1538398297-14862-1-git-send-email-magnus.karlsson@intel.com>
Previously, the xsk code did not record which umem was bound to a
specific queue id. This was not required if all drivers were zero-copy
enabled as this had to be recorded in the driver anyway. So if a user
tried to bind two umems to the same queue, the driver would say
no. But if copy-mode was first enabled and then zero-copy mode (or the
reverse order), we mistakenly enabled both of them on the same umem
leading to buggy behavior. The main culprit for this is that we did
not store the association of umem to queue id in the copy case and
only relied on the driver reporting this. As this relation was not
stored in the driver for copy mode (it does not rely on the AF_XDP
NDOs), this obviously could not work.
This patch fixes the problem by always recording the umem to queue id
relationship in the netdev_queue and netdev_rx_queue structs. This way
we always know what kind of umem has been bound to a queue id and can
act appropriately at bind time.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
net/xdp/xdp_umem.c | 90 ++++++++++++++++++++++++++++++++++++++----------------
net/xdp/xdp_umem.h | 2 +-
net/xdp/xsk.c | 7 -----
3 files changed, 64 insertions(+), 35 deletions(-)
diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c
index 555427b..4d6c665 100644
--- a/net/xdp/xdp_umem.c
+++ b/net/xdp/xdp_umem.c
@@ -42,27 +42,47 @@ void xdp_del_sk_umem(struct xdp_umem *umem, struct xdp_sock *xs)
}
}
-int xdp_umem_query(struct net_device *dev, u16 queue_id)
+/* The umem is stored both in the _rx struct and the _tx struct as we do
+ * not know if the device has more tx queues than rx, or the opposite.
+ * This might also change during run time.
+ */
+static void xdp_reg_umem_at_qid(struct net_device *dev, struct xdp_umem *umem,
+ u16 queue_id)
{
- struct netdev_bpf bpf;
+ if (queue_id < dev->real_num_rx_queues)
+ dev->_rx[queue_id].umem = umem;
+ if (queue_id < dev->real_num_tx_queues)
+ dev->_tx[queue_id].umem = umem;
+}
- ASSERT_RTNL();
+static struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev,
+ u16 queue_id)
+{
+ if (queue_id < dev->real_num_rx_queues)
+ return dev->_rx[queue_id].umem;
+ if (queue_id < dev->real_num_tx_queues)
+ return dev->_tx[queue_id].umem;
- memset(&bpf, 0, sizeof(bpf));
- bpf.command = XDP_QUERY_XSK_UMEM;
- bpf.xsk.queue_id = queue_id;
+ return NULL;
+}
- if (!dev->netdev_ops->ndo_bpf)
- return 0;
- return dev->netdev_ops->ndo_bpf(dev, &bpf) ?: !!bpf.xsk.umem;
+static void xdp_clear_umem_at_qid(struct net_device *dev, u16 queue_id)
+{
+ /* Zero out the entry independent on how many queues are configured
+ * at this point in time, as it might be used in the future.
+ */
+ if (queue_id < dev->num_rx_queues)
+ dev->_rx[queue_id].umem = NULL;
+ if (queue_id < dev->num_tx_queues)
+ dev->_tx[queue_id].umem = NULL;
}
int xdp_umem_assign_dev(struct xdp_umem *umem, struct net_device *dev,
- u32 queue_id, u16 flags)
+ u16 queue_id, u16 flags)
{
bool force_zc, force_copy;
struct netdev_bpf bpf;
- int err;
+ int err = 0;
force_zc = flags & XDP_ZEROCOPY;
force_copy = flags & XDP_COPY;
@@ -70,17 +90,23 @@ int xdp_umem_assign_dev(struct xdp_umem *umem, struct net_device *dev,
if (force_zc && force_copy)
return -EINVAL;
- if (force_copy)
- return 0;
+ rtnl_lock();
+ if (xdp_get_umem_from_qid(dev, queue_id)) {
+ err = -EBUSY;
+ goto out_rtnl_unlock;
+ }
- if (!dev->netdev_ops->ndo_bpf || !dev->netdev_ops->ndo_xsk_async_xmit)
- return force_zc ? -EOPNOTSUPP : 0; /* fail or fallback */
+ xdp_reg_umem_at_qid(dev, umem, queue_id);
+ umem->dev = dev;
+ umem->queue_id = queue_id;
+ if (force_copy)
+ /* For copy-mode, we are done. */
+ goto out_rtnl_unlock;
- rtnl_lock();
- err = xdp_umem_query(dev, queue_id);
- if (err) {
- err = err < 0 ? -EOPNOTSUPP : -EBUSY;
- goto err_rtnl_unlock;
+ if (!dev->netdev_ops->ndo_bpf ||
+ !dev->netdev_ops->ndo_xsk_async_xmit) {
+ err = -EOPNOTSUPP;
+ goto err_unreg_umem;
}
bpf.command = XDP_SETUP_XSK_UMEM;
@@ -89,18 +115,20 @@ int xdp_umem_assign_dev(struct xdp_umem *umem, struct net_device *dev,
err = dev->netdev_ops->ndo_bpf(dev, &bpf);
if (err)
- goto err_rtnl_unlock;
+ goto err_unreg_umem;
rtnl_unlock();
dev_hold(dev);
- umem->dev = dev;
- umem->queue_id = queue_id;
umem->zc = true;
return 0;
-err_rtnl_unlock:
+err_unreg_umem:
+ xdp_clear_umem_at_qid(dev, queue_id);
+ if (!force_zc)
+ err = 0; /* fallback to copy mode */
+out_rtnl_unlock:
rtnl_unlock();
- return force_zc ? err : 0; /* fail or fallback */
+ return err;
}
static void xdp_umem_clear_dev(struct xdp_umem *umem)
@@ -108,7 +136,7 @@ static void xdp_umem_clear_dev(struct xdp_umem *umem)
struct netdev_bpf bpf;
int err;
- if (umem->dev) {
+ if (umem->zc) {
bpf.command = XDP_SETUP_XSK_UMEM;
bpf.xsk.umem = NULL;
bpf.xsk.queue_id = umem->queue_id;
@@ -119,9 +147,17 @@ static void xdp_umem_clear_dev(struct xdp_umem *umem)
if (err)
WARN(1, "failed to disable umem!\n");
+ }
+
+ if (umem->dev) {
+ rtnl_lock();
+ xdp_clear_umem_at_qid(umem->dev, umem->queue_id);
+ rtnl_unlock();
+ }
+ if (umem->zc) {
dev_put(umem->dev);
- umem->dev = NULL;
+ umem->zc = false;
}
}
diff --git a/net/xdp/xdp_umem.h b/net/xdp/xdp_umem.h
index c8be1ad..2760322 100644
--- a/net/xdp/xdp_umem.h
+++ b/net/xdp/xdp_umem.h
@@ -9,7 +9,7 @@
#include <net/xdp_sock.h>
int xdp_umem_assign_dev(struct xdp_umem *umem, struct net_device *dev,
- u32 queue_id, u16 flags);
+ u16 queue_id, u16 flags);
bool xdp_umem_validate_queues(struct xdp_umem *umem);
void xdp_get_umem(struct xdp_umem *umem);
void xdp_put_umem(struct xdp_umem *umem);
diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index 5a432df..caeddad1 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c
@@ -419,13 +419,6 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
}
qid = sxdp->sxdp_queue_id;
-
- if ((xs->rx && qid >= dev->real_num_rx_queues) ||
- (xs->tx && qid >= dev->real_num_tx_queues)) {
- err = -EINVAL;
- goto out_unlock;
- }
-
flags = sxdp->sxdp_flags;
if (flags & XDP_SHARED_UMEM) {
--
2.7.4
^ permalink raw reply related
* [PATCH bpf-next v2 3/5] ethtool: rename local variable max -> curr
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
In-Reply-To: <1538398297-14862-1-git-send-email-magnus.karlsson@intel.com>
From: Jakub Kicinski <jakub.kicinski@netronome.com>
ethtool_set_channels() validates the config against driver's max
settings. It retrieves the current config and stores it in a
variable called max. This was okay when only max settings were
accessed but we will soon want to access current settings as
well, so calling the entire structure max makes the code less
readable.
While at it drop unnecessary parenthesis.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
---
net/core/ethtool.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index 96afc55..9a648fb 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -1655,7 +1655,7 @@ static noinline_for_stack int ethtool_get_channels(struct net_device *dev,
static noinline_for_stack int ethtool_set_channels(struct net_device *dev,
void __user *useraddr)
{
- struct ethtool_channels channels, max = { .cmd = ETHTOOL_GCHANNELS };
+ struct ethtool_channels channels, curr = { .cmd = ETHTOOL_GCHANNELS };
u32 max_rx_in_use = 0;
if (!dev->ethtool_ops->set_channels || !dev->ethtool_ops->get_channels)
@@ -1664,13 +1664,13 @@ static noinline_for_stack int ethtool_set_channels(struct net_device *dev,
if (copy_from_user(&channels, useraddr, sizeof(channels)))
return -EFAULT;
- dev->ethtool_ops->get_channels(dev, &max);
+ dev->ethtool_ops->get_channels(dev, &curr);
/* ensure new counts are within the maximums */
- if ((channels.rx_count > max.max_rx) ||
- (channels.tx_count > max.max_tx) ||
- (channels.combined_count > max.max_combined) ||
- (channels.other_count > max.max_other))
+ if (channels.rx_count > curr.max_rx ||
+ channels.tx_count > curr.max_tx ||
+ channels.combined_count > curr.max_combined ||
+ channels.other_count > curr.max_other)
return -EINVAL;
/* ensure the new Rx count fits within the configured Rx flow
--
2.7.4
^ permalink raw reply related
* [PATCH bpf-next v2 5/5] xsk: simplify xdp_clear_umem_at_qid implementation
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
In-Reply-To: <1538398297-14862-1-git-send-email-magnus.karlsson@intel.com>
As we now do not allow ethtool to deactivate the queue id we are
running an AF_XDP socket on, we can simplify the implementation of
xdp_clear_umem_at_qid().
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
net/xdp/xdp_umem.c | 7 ++-----
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c
index 773326f..c6007c5 100644
--- a/net/xdp/xdp_umem.c
+++ b/net/xdp/xdp_umem.c
@@ -68,12 +68,9 @@ struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev,
static void xdp_clear_umem_at_qid(struct net_device *dev, u16 queue_id)
{
- /* Zero out the entry independent on how many queues are configured
- * at this point in time, as it might be used in the future.
- */
- if (queue_id < dev->num_rx_queues)
+ if (queue_id < dev->real_num_rx_queues)
dev->_rx[queue_id].umem = NULL;
- if (queue_id < dev->num_tx_queues)
+ if (queue_id < dev->real_num_tx_queues)
dev->_tx[queue_id].umem = NULL;
}
--
2.7.4
^ permalink raw reply related
* [PATCH bpf-next v2 4/5] ethtool: don't allow disabling queues with umem installed
From: Magnus Karlsson @ 2018-10-01 12:51 UTC (permalink / raw)
To: magnus.karlsson, bjorn.topel, ast, daniel, netdev, jakub.kicinski
In-Reply-To: <1538398297-14862-1-git-send-email-magnus.karlsson@intel.com>
From: Jakub Kicinski <jakub.kicinski@netronome.com>
We already check the RSS indirection table does not use queues which
would be disabled by channel reconfiguration. Make sure user does not
try to disable queues which have a UMEM and zero-copy AF_XDP socket
installed.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
---
include/net/xdp_sock.h | 7 +++++++
net/core/ethtool.c | 11 +++++++++++
net/xdp/xdp_umem.c | 4 ++--
3 files changed, 20 insertions(+), 2 deletions(-)
diff --git a/include/net/xdp_sock.h b/include/net/xdp_sock.h
index 70a115b..13acb98 100644
--- a/include/net/xdp_sock.h
+++ b/include/net/xdp_sock.h
@@ -86,6 +86,7 @@ struct xdp_umem_fq_reuse *xsk_reuseq_prepare(u32 nentries);
struct xdp_umem_fq_reuse *xsk_reuseq_swap(struct xdp_umem *umem,
struct xdp_umem_fq_reuse *newq);
void xsk_reuseq_free(struct xdp_umem_fq_reuse *rq);
+struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev, u16 queue_id);
static inline char *xdp_umem_get_data(struct xdp_umem *umem, u64 addr)
{
@@ -183,6 +184,12 @@ static inline void xsk_reuseq_free(struct xdp_umem_fq_reuse *rq)
{
}
+static inline struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev,
+ u16 queue_id)
+{
+ return NULL;
+}
+
static inline char *xdp_umem_get_data(struct xdp_umem *umem, u64 addr)
{
return NULL;
diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index 9a648fb..5a788ad 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -27,6 +27,7 @@
#include <linux/rtnetlink.h>
#include <linux/sched/signal.h>
#include <linux/net.h>
+#include <net/xdp_sock.h>
/*
* Some useful ethtool_ops methods that're device independent.
@@ -1656,7 +1657,9 @@ static noinline_for_stack int ethtool_set_channels(struct net_device *dev,
void __user *useraddr)
{
struct ethtool_channels channels, curr = { .cmd = ETHTOOL_GCHANNELS };
+ u16 from_channel, to_channel;
u32 max_rx_in_use = 0;
+ unsigned int i;
if (!dev->ethtool_ops->set_channels || !dev->ethtool_ops->get_channels)
return -EOPNOTSUPP;
@@ -1680,6 +1683,14 @@ static noinline_for_stack int ethtool_set_channels(struct net_device *dev,
(channels.combined_count + channels.rx_count) <= max_rx_in_use)
return -EINVAL;
+ /* Disabling channels, query zero-copy AF_XDP sockets */
+ from_channel = channels.combined_count +
+ min(channels.rx_count, channels.tx_count);
+ to_channel = curr.combined_count + max(curr.rx_count, curr.tx_count);
+ for (i = from_channel; i < to_channel; i++)
+ if (xdp_get_umem_from_qid(dev, i))
+ return -EINVAL;
+
return dev->ethtool_ops->set_channels(dev, &channels);
}
diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c
index 4d6c665..773326f 100644
--- a/net/xdp/xdp_umem.c
+++ b/net/xdp/xdp_umem.c
@@ -55,8 +55,8 @@ static void xdp_reg_umem_at_qid(struct net_device *dev, struct xdp_umem *umem,
dev->_tx[queue_id].umem = umem;
}
-static struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev,
- u16 queue_id)
+struct xdp_umem *xdp_get_umem_from_qid(struct net_device *dev,
+ u16 queue_id)
{
if (queue_id < dev->real_num_rx_queues)
return dev->_rx[queue_id].umem;
--
2.7.4
^ permalink raw reply related
* Re: [PATCH 1/1] macsec: reflect the master interface state
From: Sabrina Dubroca @ 2018-10-01 12:52 UTC (permalink / raw)
To: Radu Rendec; +Cc: netdev, davem, ptalbert
In-Reply-To: <CAD5jUk8gpiX_AyiRPf7fnBouyxWS7V8FEAiiug1A3iMdr_YJBA@mail.gmail.com>
2018-09-19, 12:44:41 -0400, Radu Rendec wrote:
> Hello,
Gah, sorry, this fell into the "week-end" crack and I forgot to answer :(
> On Wed, Sep 19, 2018 at 11:24 AM Sabrina Dubroca <sd@queasysnail.net> wrote:
> > 2018-09-18, 20:16:12 -0400, Radu Rendec wrote:
> > > This patch makes macsec interfaces reflect the state of the underlying
> > > interface: if the master interface changes state to up/down, the macsec
> > > interface changes its state accordingly.
> >
> > Yes, that's a bit unfortunate. I was wondering if we should just allow
> > setting the device up, and let link state handle the rest.
>
> Yes, that could work. It would also be consistent with the idea of
> propagating only the link state.
Do you want to handle it, or should I?
> > It's missing small parts of link state handling that I have been
> > testing, mainly when creating a new device.
>
> Can you please be more specific? I've just looked at macsec_newlink()
> and I didn't notice anything related to link state handling.
Yes, that's actually the problem. For example,
macvlan_common_newlink() does:
netif_stacked_transfer_operstate(lowerdev, dev);
linkwatch_fire_event(dev);
If you try to create a macsec device UP with its lowerdev UP:
ip link set $lower up
ip link add link $lower up type macsec
You'll get:
macsec0@$lower: <BROADCAST,MULTICAST,UP,LOWER_UP> [...] state UNKNOWN [...]
and you want "state UP".
> > My version of the patch only does netif_stacked_transfer_operstate(),
> > and skips setting the device administratively down (ie, same as
> > NETDEV_CHANGE).
>
> That makes sense. But, since you mentioned you had your own patch, does
> it make sense for me to continue working on mine?
Here's what I have (without a commit message, because I hadn't written
one yet -- yours looks pretty good, other than missing a "Fixes:" tag):
diff --git a/drivers/net/macsec.c b/drivers/net/macsec.c
index 7de88b33d5b9..3532cdee2761 100644
--- a/drivers/net/macsec.c
+++ b/drivers/net/macsec.c
@@ -3308,6 +3308,9 @@ static int macsec_newlink(struct net *net, struct net_device *dev,
if (err < 0)
goto del_dev;
+ netif_stacked_transfer_operstate(real_dev, dev);
+ linkwatch_fire_event(dev);
+
macsec_generation++;
return 0;
@@ -3492,6 +3495,20 @@ static int macsec_notify(struct notifier_block *this, unsigned long event,
return NOTIFY_DONE;
switch (event) {
+ case NETDEV_DOWN:
+ case NETDEV_UP:
+ case NETDEV_CHANGE: {
+ struct macsec_dev *m, *n;
+ struct macsec_rxh_data *rxd;
+
+ rxd = macsec_data_rtnl(real_dev);
+ list_for_each_entry_safe(m, n, &rxd->secys, secys) {
+ struct net_device *dev = m->secy.netdev;
+
+ netif_stacked_transfer_operstate(real_dev, dev);
+ }
+ break;
+ }
case NETDEV_UNREGISTER: {
struct macsec_dev *m, *n;
struct macsec_rxh_data *rxd;
If you want to keep working on this, that's okay for me. I'm not hung
up on who gets authorship.
> > > }
> > > + case NETDEV_UP:
> > > + list_for_each_entry(m, &rxd->secys, secys) {
> > > + struct net_device *dev = m->secy.netdev;
> > > + int flags = dev_get_flags(dev);
> > > +
> > > + if (!(flags & IFF_UP)) {
> > > + dev_change_flags(dev, flags | IFF_UP);
> > > + netif_stacked_transfer_operstate(real_dev, dev);
> > > + }
> > > + }
> > > + break;
> >
> > I don't like that this completely ignores whether the device was done
> > independently of the lower link. Maybe the administrator actually
> > wants the macsec device down, regardless of state changes on the lower
> > device.
>
> I thought about that too and I don't like it either. Perhaps the vlan
> approach of having a "loose binding" flag is worth considering. Then the
> admin can decice for themselves.
Do you have a use case where you'd want that? If that solves some
problem for you (a problem that can't be solved just by fixing the
current bug), then ok, we can consider it. I prefer to avoid adding
obscure options and more code unless they're necessary.
> However, I guess the problem disappears if only the link state is
> propagated. The only caveat to that is to not propagate an "up" link
> state while the macsec interface is administratively down, but
> "remember" to propagate it later if the macsec interface is
> administratively set to "up".
>
> Thanks for the feedback!
And sorry for the delay in answering :/
--
Sabrina
^ permalink raw reply related
* Re: [PATCH] dt-bindings: can: rcar_can: Add r8a7744 support
From: Simon Horman @ 2018-10-01 12:56 UTC (permalink / raw)
To: Biju Das
Cc: Wolfgang Grandegger, Marc Kleine-Budde, Rob Herring, Mark Rutland,
David S. Miller, linux-can, netdev, devicetree,
Geert Uytterhoeven, Chris Paterson, Fabrizio Castro,
linux-renesas-soc
In-Reply-To: <1538050128-47356-1-git-send-email-biju.das@bp.renesas.com>
On Thu, Sep 27, 2018 at 01:08:48PM +0100, Biju Das wrote:
> Document RZ/G1N (r8a7744) SoC specific bindings.
>
> Signed-off-by: Biju Das <biju.das@bp.renesas.com>
> Reviewed-by: Chris Paterson <Chris.Paterson2@renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
^ permalink raw reply
* [PATCH net] rhashtable: prevent work queue schedule while dismantling
From: Eric Dumazet @ 2018-10-01 13:16 UTC (permalink / raw)
To: David S . Miller
Cc: netdev, Eric Dumazet, Eric Dumazet, Thomas Graf, Herbert Xu
syszbot found an interesting use-after-free [1] happening
while IPv4 fragment rhashtable was destroyed at netns dismantle.
While no insertions can possibly happen at the time a dismantling
netns is destroying this rhashtable, timers can still fire and
attempt to remove elements from this rhashtable.
Since automatic shrinking is enabled, this might then schedule
the work queue to perform the shrinking from process context,
which was not expected.
I have thought of simply forcing ht->p.automatic_shrinking to
false before the cancel_work_sync(&ht->run_work);
in rhashtable_free_and_destroy(), but that would be still racy.
Lets use a separate boolean to not only prevent the work queue
being scheduled, but also make sure rht_deferred_worker() would
return early.
[1]
BUG: KASAN: use-after-free in __read_once_size include/linux/compiler.h:188 [inline]
BUG: KASAN: use-after-free in rhashtable_last_table+0x216/0x240 lib/rhashtable.c:217
Read of size 8 at addr ffff88019a4c8840 by task kworker/0:4/8279
CPU: 0 PID: 8279 Comm: kworker/0:4 Not tainted 4.19.0-rc5+ #61
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: events rht_deferred_worker
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x1c4/0x2b4 lib/dump_stack.c:113
print_address_description.cold.8+0x9/0x1ff mm/kasan/report.c:256
kasan_report_error mm/kasan/report.c:354 [inline]
kasan_report.cold.9+0x242/0x309 mm/kasan/report.c:412
__asan_report_load8_noabort+0x14/0x20 mm/kasan/report.c:433
__read_once_size include/linux/compiler.h:188 [inline]
rhashtable_last_table+0x216/0x240 lib/rhashtable.c:217
rht_deferred_worker+0x157/0x1de0 lib/rhashtable.c:410
process_one_work+0xc90/0x1b90 kernel/workqueue.c:2153
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413
Allocated by task 5:
save_stack+0x43/0xd0 mm/kasan/kasan.c:448
set_track mm/kasan/kasan.c:460 [inline]
kasan_kmalloc+0xc7/0xe0 mm/kasan/kasan.c:553
__do_kmalloc_node mm/slab.c:3682 [inline]
__kmalloc_node+0x47/0x70 mm/slab.c:3689
kmalloc_node include/linux/slab.h:555 [inline]
kvmalloc_node+0xb9/0xf0 mm/util.c:423
kvmalloc include/linux/mm.h:577 [inline]
kvzalloc include/linux/mm.h:585 [inline]
bucket_table_alloc+0x9a/0x4e0 lib/rhashtable.c:176
rhashtable_rehash_alloc+0x73/0x100 lib/rhashtable.c:353
rht_deferred_worker+0x278/0x1de0 lib/rhashtable.c:413
process_one_work+0xc90/0x1b90 kernel/workqueue.c:2153
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413
Freed by task 8283:
save_stack+0x43/0xd0 mm/kasan/kasan.c:448
set_track mm/kasan/kasan.c:460 [inline]
__kasan_slab_free+0x102/0x150 mm/kasan/kasan.c:521
kasan_slab_free+0xe/0x10 mm/kasan/kasan.c:528
__cache_free mm/slab.c:3498 [inline]
kfree+0xcf/0x230 mm/slab.c:3813
kvfree+0x61/0x70 mm/util.c:452
bucket_table_free+0xda/0x250 lib/rhashtable.c:108
rhashtable_free_and_destroy+0x152/0x900 lib/rhashtable.c:1163
inet_frags_exit_net+0x3d/0x50 net/ipv4/inet_fragment.c:96
ipv4_frags_exit_net+0x73/0x90 net/ipv4/ip_fragment.c:914
ops_exit_list.isra.7+0xb0/0x160 net/core/net_namespace.c:153
cleanup_net+0x555/0xb10 net/core/net_namespace.c:551
process_one_work+0xc90/0x1b90 kernel/workqueue.c:2153
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413
The buggy address belongs to the object at ffff88019a4c8800
which belongs to the cache kmalloc-16384 of size 16384
The buggy address is located 64 bytes inside of
16384-byte region [ffff88019a4c8800, ffff88019a4cc800)
The buggy address belongs to the page:
page:ffffea0006693200 count:1 mapcount:0 mapping:ffff8801da802200 index:0x0 compound_mapcount: 0
flags: 0x2fffc0000008100(slab|head)
raw: 02fffc0000008100 ffffea0006685608 ffffea0006617c08 ffff8801da802200
raw: 0000000000000000 ffff88019a4c8800 0000000100000001 0000000000000000
page dumped because: kasan: bad access detected
Memory state around the buggy address:
ffff88019a4c8700: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
ffff88019a4c8780: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>ffff88019a4c8800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff88019a4c8880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff88019a4c8900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Thomas Graf <tgraf@suug.ch>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
---
include/linux/rhashtable-types.h | 2 ++
include/linux/rhashtable.h | 1 +
lib/rhashtable.c | 25 ++++++++++++++-----------
3 files changed, 17 insertions(+), 11 deletions(-)
diff --git a/include/linux/rhashtable-types.h b/include/linux/rhashtable-types.h
index 763d613ce2c2f275037c3b1b5ce4d390394295c6..03d86e79dfd664c48c5e437c609deb62cbc33180 100644
--- a/include/linux/rhashtable-types.h
+++ b/include/linux/rhashtable-types.h
@@ -75,6 +75,7 @@ struct rhashtable_params {
* @max_elems: Maximum number of elements in table
* @p: Configuration parameters
* @rhlist: True if this is an rhltable
+ * @dead: True if in dismantle phase
* @run_work: Deferred worker to expand/shrink asynchronously
* @mutex: Mutex to protect current/future table swapping
* @lock: Spin lock to protect walker list
@@ -86,6 +87,7 @@ struct rhashtable {
unsigned int max_elems;
struct rhashtable_params p;
bool rhlist;
+ bool dead;
struct work_struct run_work;
struct mutex mutex;
spinlock_t lock;
diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h
index eb71110392479784db1d4a0b9d86d2eee6631789..5efc7d5cc66ba37c70c7e3d5ff98473df6a3837e 100644
--- a/include/linux/rhashtable.h
+++ b/include/linux/rhashtable.h
@@ -934,6 +934,7 @@ static inline int __rhashtable_remove_fast_one(
if (err > 0) {
atomic_dec(&ht->nelems);
if (unlikely(ht->p.automatic_shrinking &&
+ !ht->dead &&
rht_shrink_below_30(ht, tbl)))
schedule_work(&ht->run_work);
err = 0;
diff --git a/lib/rhashtable.c b/lib/rhashtable.c
index 30526afa8343124f06f0649592ee246dc4d88fbe..99f9aa5a768d849d9afe3f50f25c300c37d9c07d 100644
--- a/lib/rhashtable.c
+++ b/lib/rhashtable.c
@@ -406,19 +406,21 @@ static void rht_deferred_worker(struct work_struct *work)
ht = container_of(work, struct rhashtable, run_work);
mutex_lock(&ht->mutex);
- tbl = rht_dereference(ht->tbl, ht);
- tbl = rhashtable_last_table(ht, tbl);
+ if (!ht->dead) {
+ tbl = rht_dereference(ht->tbl, ht);
+ tbl = rhashtable_last_table(ht, tbl);
- if (rht_grow_above_75(ht, tbl))
- err = rhashtable_rehash_alloc(ht, tbl, tbl->size * 2);
- else if (ht->p.automatic_shrinking && rht_shrink_below_30(ht, tbl))
- err = rhashtable_shrink(ht);
- else if (tbl->nest)
- err = rhashtable_rehash_alloc(ht, tbl, tbl->size);
-
- if (!err)
- err = rhashtable_rehash_table(ht);
+ if (rht_grow_above_75(ht, tbl))
+ err = rhashtable_rehash_alloc(ht, tbl, tbl->size * 2);
+ else if (ht->p.automatic_shrinking &&
+ rht_shrink_below_30(ht, tbl))
+ err = rhashtable_shrink(ht);
+ else if (tbl->nest)
+ err = rhashtable_rehash_alloc(ht, tbl, tbl->size);
+ if (!err)
+ err = rhashtable_rehash_table(ht);
+ }
mutex_unlock(&ht->mutex);
if (err)
@@ -1138,6 +1140,7 @@ void rhashtable_free_and_destroy(struct rhashtable *ht,
struct bucket_table *tbl, *next_tbl;
unsigned int i;
+ ht->dead = true;
cancel_work_sync(&ht->run_work);
mutex_lock(&ht->mutex);
--
2.19.0.605.g01d371f741-goog
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox