* Re: linux-next: zillions of lockdep whinges in include/net/sock.h:1408
From: David Miller @ 2016-04-24 18:54 UTC (permalink / raw)
To: hannes; +Cc: eric.dumazet, Valdis.Kletnieks, netdev, linux-kernel
In-Reply-To: <571D14F8.6070306@stressinduktion.org>
From: Hannes Frederic Sowa <hannes@stressinduktion.org>
Date: Sun, 24 Apr 2016 20:48:24 +0200
> Eric's patch is worth to apply anyway, but I am not sure if it solves
> the (fundamental) problem. I couldn't reproduce it with the exact next-
> tag provided in the initial mail. All other reports also only happend
> with linux-next and not net-next.
Ok, Eric please submit it formally.
Thanks!
^ permalink raw reply
* [PATCH net-next 0/6] net: core: remove TX_LOCKED support
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel
Not that many users left, lets kill it.
TX_LOCKED was meant to be used by LLTX drivers when spin_trylock()
failed. Stack then re-queued if collisions happened on different
cpus or free'd the skb to prevent deadlocks.
Most of the driver removal patches fall into one of three categories:
1. remove the driver-private tx lock (and LLTX flag), or...
2. convert spin_trylock to plain spin_lock, or...
3. convert TX_LOCKED to free+TX_OK
Patches are grouped by these categories, last patch is the actual removal.
All driver changes were compile tested only with exception of atl1e.
Documentation/networking/netdev-features.txt | 10 ++---
Documentation/networking/netdevices.txt | 9 +----
drivers/infiniband/hw/nes/nes_nic.c | 13 ++-----
drivers/net/ethernet/amd/7990.c | 8 ++--
drivers/net/ethernet/amd/a2065.c | 7 +---
drivers/net/ethernet/atheros/atl1c/atl1c.h | 3 -
drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 11 ------
drivers/net/ethernet/atheros/atl1e/atl1e.h | 1
drivers/net/ethernet/atheros/atl1e/atl1e_main.c | 12 -------
drivers/net/ethernet/chelsio/cxgb/sge.c | 3 -
drivers/net/ethernet/dec/tulip/de4x5.c | 7 ++--
drivers/net/ethernet/neterion/s2io.c | 9 -----
drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c | 6 +--
drivers/net/ethernet/tehuti/tehuti.c | 8 ----
drivers/net/hamradio/baycom_epp.c | 6 ++-
drivers/net/hamradio/hdlcdrv.c | 6 ++-
drivers/net/rionet.c | 6 ---
include/linux/netdevice.h | 3 -
net/core/net-procfs.c | 3 +
net/core/pktgen.c | 1
net/sched/sch_generic.c | 32 -------------------
21 files changed, 43 insertions(+), 121 deletions(-)
^ permalink raw reply
* [PATCH net-next 2/6] atl1c: remove private tx lock
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel, Florian Westphal, Jay Cliburn, Chris Snook
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
AFAICS this is safe: the lock is only used in the .ndo_start_xmit
function and this driver does not set LLTX.
Gets rid of TX_LOCKED return value, followup patches will remove it.
Cc: Jay Cliburn <jcliburn@gmail.com>
Cc: Chris Snook <chris.snook@gmail.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
drivers/net/ethernet/atheros/atl1c/atl1c.h | 3 +--
drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 11 -----------
2 files changed, 1 insertion(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c.h b/drivers/net/ethernet/atheros/atl1c/atl1c.h
index b9203d9..c46b489 100644
--- a/drivers/net/ethernet/atheros/atl1c/atl1c.h
+++ b/drivers/net/ethernet/atheros/atl1c/atl1c.h
@@ -488,7 +488,7 @@ struct atl1c_tpd_ring {
dma_addr_t dma; /* descriptor ring physical address */
u16 size; /* descriptor ring length in bytes */
u16 count; /* number of descriptors in the ring */
- u16 next_to_use; /* this is protectd by adapter->tx_lock */
+ u16 next_to_use;
atomic_t next_to_clean;
struct atl1c_buffer *buffer_info;
};
@@ -542,7 +542,6 @@ struct atl1c_adapter {
u16 link_duplex;
spinlock_t mdio_lock;
- spinlock_t tx_lock;
atomic_t irq_sem;
struct work_struct common_task;
diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
index d0084d4..a3200ea 100644
--- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
+++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c
@@ -821,7 +821,6 @@ static int atl1c_sw_init(struct atl1c_adapter *adapter)
atl1c_set_rxbufsize(adapter, adapter->netdev);
atomic_set(&adapter->irq_sem, 1);
spin_lock_init(&adapter->mdio_lock);
- spin_lock_init(&adapter->tx_lock);
set_bit(__AT_DOWN, &adapter->flags);
return 0;
@@ -2206,7 +2205,6 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb,
struct net_device *netdev)
{
struct atl1c_adapter *adapter = netdev_priv(netdev);
- unsigned long flags;
u16 tpd_req = 1;
struct atl1c_tpd_desc *tpd;
enum atl1c_trans_queue type = atl1c_trans_normal;
@@ -2217,16 +2215,10 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb,
}
tpd_req = atl1c_cal_tpd_req(skb);
- if (!spin_trylock_irqsave(&adapter->tx_lock, flags)) {
- if (netif_msg_pktdata(adapter))
- dev_info(&adapter->pdev->dev, "tx locked\n");
- return NETDEV_TX_LOCKED;
- }
if (atl1c_tpd_avail(adapter, type) < tpd_req) {
/* no enough descriptor, just stop queue */
netif_stop_queue(netdev);
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
return NETDEV_TX_BUSY;
}
@@ -2234,7 +2226,6 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb,
/* do TSO and check sum */
if (atl1c_tso_csum(adapter, skb, &tpd, type) != 0) {
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
}
@@ -2257,12 +2248,10 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb,
"tx-skb droppted due to dma error\n");
/* roll back tpd/buffer */
atl1c_tx_rollback(adapter, tpd, type);
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
dev_kfree_skb_any(skb);
} else {
netdev_sent_queue(adapter->netdev, skb->len);
atl1c_tx_queue(adapter, skb, tpd, type);
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
}
return NETDEV_TX_OK;
--
2.7.3
^ permalink raw reply related
* [PATCH net-next 3/6] atle1: remove LLTX support and TX_UNLOCKED
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel, Florian Westphal, Jay Cliburn, Chris Snook
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
similar to atl1c: lock is only used in ndo_start_xmit, but we also
advertised LLTX, so remove that as well and let core stack handle
tx locking.
Allows to remove the TX_LOCKED return value from the driver.
Cc: Jay Cliburn <jcliburn@gmail.com>
Cc: Chris Snook <chris.snook@gmail.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
drivers/net/ethernet/atheros/atl1e/atl1e.h | 1 -
drivers/net/ethernet/atheros/atl1e/atl1e_main.c | 12 +-----------
2 files changed, 1 insertion(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e.h b/drivers/net/ethernet/atheros/atl1e/atl1e.h
index 0212dac..632bb84 100644
--- a/drivers/net/ethernet/atheros/atl1e/atl1e.h
+++ b/drivers/net/ethernet/atheros/atl1e/atl1e.h
@@ -442,7 +442,6 @@ struct atl1e_adapter {
u16 link_duplex;
spinlock_t mdio_lock;
- spinlock_t tx_lock;
atomic_t irq_sem;
struct work_struct reset_task;
diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
index 59a03a1..974713b 100644
--- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
+++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c
@@ -648,7 +648,6 @@ static int atl1e_sw_init(struct atl1e_adapter *adapter)
atomic_set(&adapter->irq_sem, 1);
spin_lock_init(&adapter->mdio_lock);
- spin_lock_init(&adapter->tx_lock);
set_bit(__AT_DOWN, &adapter->flags);
@@ -1866,7 +1865,6 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb,
struct net_device *netdev)
{
struct atl1e_adapter *adapter = netdev_priv(netdev);
- unsigned long flags;
u16 tpd_req = 1;
struct atl1e_tpd_desc *tpd;
@@ -1880,13 +1878,10 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb,
return NETDEV_TX_OK;
}
tpd_req = atl1e_cal_tdp_req(skb);
- if (!spin_trylock_irqsave(&adapter->tx_lock, flags))
- return NETDEV_TX_LOCKED;
if (atl1e_tpd_avail(adapter) < tpd_req) {
/* no enough descriptor, just stop queue */
netif_stop_queue(netdev);
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
return NETDEV_TX_BUSY;
}
@@ -1910,7 +1905,6 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb,
/* do TSO and check sum */
if (atl1e_tso_csum(adapter, skb, tpd) != 0) {
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
}
@@ -1921,10 +1915,7 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb,
}
atl1e_tx_queue(adapter, tpd_req, tpd);
-
- netdev->trans_start = jiffies; /* NETIF_F_LLTX driver :( */
out:
- spin_unlock_irqrestore(&adapter->tx_lock, flags);
return NETDEV_TX_OK;
}
@@ -2285,8 +2276,7 @@ static int atl1e_init_netdev(struct net_device *netdev, struct pci_dev *pdev)
netdev->hw_features = NETIF_F_SG | NETIF_F_HW_CSUM | NETIF_F_TSO |
NETIF_F_HW_VLAN_CTAG_RX;
- netdev->features = netdev->hw_features | NETIF_F_LLTX |
- NETIF_F_HW_VLAN_CTAG_TX;
+ netdev->features = netdev->hw_features | NETIF_F_HW_VLAN_CTAG_TX;
/* not enabled by default */
netdev->hw_features |= NETIF_F_RXALL | NETIF_F_RXFCS;
return 0;
--
2.7.3
^ permalink raw reply related
* [PATCH net-next 4/6] drivers: net: use NETDEV_TX_OK instead of NETDEV_TX_LOCKED
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, Florian Westphal, linux-parisc, linux-hams,
Thomas Sailer
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
These drivers already call netif_stop_queue() so we should not be called
unless tx space is available. Just free the skb and return TX_OK.
Followup patch will remove NETDEV_TX_LOCKED from the kernel.
Cc: linux-parisc@vger.kernel.org
Cc: linux-hams@vger.kernel.org
Cc: Thomas Sailer <t.sailer@alumni.ethz.ch>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
checkpatch chokes on this one, tulip uses spaces, not tabs...
drivers/net/ethernet/amd/7990.c | 8 +++++---
drivers/net/ethernet/amd/a2065.c | 7 +++----
drivers/net/ethernet/dec/tulip/de4x5.c | 7 +++++--
drivers/net/hamradio/baycom_epp.c | 6 ++++--
drivers/net/hamradio/hdlcdrv.c | 6 ++++--
5 files changed, 21 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/amd/7990.c b/drivers/net/ethernet/amd/7990.c
index 66d0b73c..8e75755 100644
--- a/drivers/net/ethernet/amd/7990.c
+++ b/drivers/net/ethernet/amd/7990.c
@@ -543,11 +543,13 @@ int lance_start_xmit(struct sk_buff *skb, struct net_device *dev)
static int outs;
unsigned long flags;
- if (!TX_BUFFS_AVAIL)
- return NETDEV_TX_LOCKED;
-
netif_stop_queue(dev);
+ if (!TX_BUFFS_AVAIL) {
+ dev_consume_skb_any(skb);
+ return NETDEV_TX_OK;
+ }
+
skblen = skb->len;
#ifdef DEBUG_DRIVER
diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c
index 5613918..2a18d34 100644
--- a/drivers/net/ethernet/amd/a2065.c
+++ b/drivers/net/ethernet/amd/a2065.c
@@ -547,10 +547,8 @@ static netdev_tx_t lance_start_xmit(struct sk_buff *skb,
local_irq_save(flags);
- if (!lance_tx_buffs_avail(lp)) {
- local_irq_restore(flags);
- return NETDEV_TX_LOCKED;
- }
+ if (!lance_tx_buffs_avail(lp))
+ goto out_free;
#ifdef DEBUG
/* dump the packet */
@@ -573,6 +571,7 @@ static netdev_tx_t lance_start_xmit(struct sk_buff *skb,
/* Kick the lance: transmit now */
ll->rdp = LE_C0_INEA | LE_C0_TDMD;
+ out_free:
dev_kfree_skb(skb);
local_irq_restore(flags);
diff --git a/drivers/net/ethernet/dec/tulip/de4x5.c b/drivers/net/ethernet/dec/tulip/de4x5.c
index 3acde3b..d88fbab 100644
--- a/drivers/net/ethernet/dec/tulip/de4x5.c
+++ b/drivers/net/ethernet/dec/tulip/de4x5.c
@@ -1465,7 +1465,7 @@ de4x5_queue_pkt(struct sk_buff *skb, struct net_device *dev)
netif_stop_queue(dev);
if (!lp->tx_enable) /* Cannot send for now */
- return NETDEV_TX_LOCKED;
+ goto tx_err;
/*
** Clean out the TX ring asynchronously to interrupts - sometimes the
@@ -1478,7 +1478,7 @@ de4x5_queue_pkt(struct sk_buff *skb, struct net_device *dev)
/* Test if cache is already locked - requeue skb if so */
if (test_and_set_bit(0, (void *)&lp->cache.lock) && !lp->interrupt)
- return NETDEV_TX_LOCKED;
+ goto tx_err;
/* Transmit descriptor ring full or stale skb */
if (netif_queue_stopped(dev) || (u_long) lp->tx_skb[lp->tx_new] > 1) {
@@ -1519,6 +1519,9 @@ de4x5_queue_pkt(struct sk_buff *skb, struct net_device *dev)
lp->cache.lock = 0;
return NETDEV_TX_OK;
+tx_err:
+ dev_kfree_skb_any(skb);
+ return NETDEV_TX_OK;
}
/*
diff --git a/drivers/net/hamradio/baycom_epp.c b/drivers/net/hamradio/baycom_epp.c
index 72c9f1f..eb66638 100644
--- a/drivers/net/hamradio/baycom_epp.c
+++ b/drivers/net/hamradio/baycom_epp.c
@@ -780,8 +780,10 @@ static int baycom_send_packet(struct sk_buff *skb, struct net_device *dev)
dev_kfree_skb(skb);
return NETDEV_TX_OK;
}
- if (bc->skb)
- return NETDEV_TX_LOCKED;
+ if (bc->skb) {
+ dev_kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
/* strip KISS byte */
if (skb->len >= HDLCDRV_MAXFLEN+1 || skb->len < 3) {
dev_kfree_skb(skb);
diff --git a/drivers/net/hamradio/hdlcdrv.c b/drivers/net/hamradio/hdlcdrv.c
index 49fe59b..4bad0b8 100644
--- a/drivers/net/hamradio/hdlcdrv.c
+++ b/drivers/net/hamradio/hdlcdrv.c
@@ -412,8 +412,10 @@ static netdev_tx_t hdlcdrv_send_packet(struct sk_buff *skb,
dev_kfree_skb(skb);
return NETDEV_TX_OK;
}
- if (sm->skb)
- return NETDEV_TX_LOCKED;
+ if (sm->skb) {
+ dev_kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
netif_stop_queue(dev);
sm->skb = skb;
return NETDEV_TX_OK;
--
2.7.3
^ permalink raw reply related
* [PATCH net-next 1/6] RDMA/nes: remove use of NETDEV_TX_LOCKED
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel, Florian Westphal, linux-rdma
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
ndo_start_xmit never returns it to stack, but nes_nic_send helper used it if
skb could not be queued to hardware. Switch to bool instead.
Cc: <linux-rdma@vger.kernel.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
drivers/infiniband/hw/nes/nes_nic.c | 13 +++++--------
1 file changed, 5 insertions(+), 8 deletions(-)
diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 3ea9e05..b09a6db 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -356,7 +356,7 @@ static int nes_netdev_stop(struct net_device *netdev)
/**
* nes_nic_send
*/
-static int nes_nic_send(struct sk_buff *skb, struct net_device *netdev)
+static bool nes_nic_send(struct sk_buff *skb, struct net_device *netdev)
{
struct nes_vnic *nesvnic = netdev_priv(netdev);
struct nes_device *nesdev = nesvnic->nesdev;
@@ -413,7 +413,7 @@ static int nes_nic_send(struct sk_buff *skb, struct net_device *netdev)
netdev->name, skb_shinfo(skb)->nr_frags + 2, skb_headlen(skb));
kfree_skb(skb);
nesvnic->tx_sw_dropped++;
- return NETDEV_TX_LOCKED;
+ return false;
}
set_bit(nesnic->sq_head, nesnic->first_frag_overflow);
bus_address = pci_map_single(nesdev->pcidev, skb->data + NES_FIRST_FRAG_SIZE,
@@ -454,8 +454,7 @@ static int nes_nic_send(struct sk_buff *skb, struct net_device *netdev)
set_wqe_32bit_value(nic_sqe->wqe_words, NES_NIC_SQ_WQE_MISC_IDX, wqe_misc);
nesnic->sq_head++;
nesnic->sq_head &= nesnic->sq_size - 1;
-
- return NETDEV_TX_OK;
+ return true;
}
@@ -673,13 +672,11 @@ tso_sq_no_longer_full:
skb_linearize(skb);
skb_set_transport_header(skb, hoffset);
skb_set_network_header(skb, nhoffset);
- send_rc = nes_nic_send(skb, netdev);
- if (send_rc != NETDEV_TX_OK)
+ if (!nes_nic_send(skb, netdev))
return NETDEV_TX_OK;
}
} else {
- send_rc = nes_nic_send(skb, netdev);
- if (send_rc != NETDEV_TX_OK)
+ if (!nes_nic_send(skb, netdev))
return NETDEV_TX_OK;
}
--
2.7.3
^ permalink raw reply related
* [PATCH net-next 6/6] net: remove NETDEV_TX_LOCKED support
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel, Florian Westphal
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
No more users in the tree, remove NETDEV_TX_LOCKED support.
Adds another hole in softnet_stats struct, but better than keeping
the unused collision counter around.
Signed-off-by: Florian Westphal <fw@strlen.de>
---
Documentation/networking/netdev-features.txt | 10 ++++-----
Documentation/networking/netdevices.txt | 9 +++-----
include/linux/netdevice.h | 3 ---
net/core/net-procfs.c | 3 ++-
net/core/pktgen.c | 1 -
net/sched/sch_generic.c | 32 ----------------------------
6 files changed, 9 insertions(+), 49 deletions(-)
diff --git a/Documentation/networking/netdev-features.txt b/Documentation/networking/netdev-features.txt
index f310ede..7413eb0 100644
--- a/Documentation/networking/netdev-features.txt
+++ b/Documentation/networking/netdev-features.txt
@@ -131,13 +131,11 @@ stack. Driver should not change behaviour based on them.
* LLTX driver (deprecated for hardware drivers)
-NETIF_F_LLTX should be set in drivers that implement their own locking in
-transmit path or don't need locking at all (e.g. software tunnels).
-In ndo_start_xmit, it is recommended to use a try_lock and return
-NETDEV_TX_LOCKED when the spin lock fails. The locking should also properly
-protect against other callbacks (the rules you need to find out).
+NETIF_F_LLTX is meant to be used by drivers that don't need locking at all,
+e.g. software tunnels.
-Don't use it for new drivers.
+This is also used in a few legacy drivers that implement their
+own locking, don't use it for new (hardware) drivers.
* netns-local device
diff --git a/Documentation/networking/netdevices.txt b/Documentation/networking/netdevices.txt
index 0b1cf6b..7fec206 100644
--- a/Documentation/networking/netdevices.txt
+++ b/Documentation/networking/netdevices.txt
@@ -69,10 +69,9 @@ ndo_start_xmit:
When the driver sets NETIF_F_LLTX in dev->features this will be
called without holding netif_tx_lock. In this case the driver
- has to lock by itself when needed. It is recommended to use a try lock
- for this and return NETDEV_TX_LOCKED when the spin lock fails.
- The locking there should also properly protect against
- set_rx_mode. Note that the use of NETIF_F_LLTX is deprecated.
+ has to lock by itself when needed.
+ The locking there should also properly protect against
+ set_rx_mode. WARNING: use of NETIF_F_LLTX is deprecated.
Don't use it for new drivers.
Context: Process with BHs disabled or BH (timer),
@@ -83,8 +82,6 @@ ndo_start_xmit:
o NETDEV_TX_BUSY Cannot transmit packet, try later
Usually a bug, means queue start/stop flow control is broken in
the driver. Note: the driver must NOT put the skb in its DMA ring.
- o NETDEV_TX_LOCKED Locking failed, please retry quickly.
- Only valid when NETIF_F_LLTX is set.
ndo_tx_timeout:
Synchronization: netif_tx_lock spinlock; all TX queues frozen.
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 1f6d5db..18d8394 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -106,7 +106,6 @@ enum netdev_tx {
__NETDEV_TX_MIN = INT_MIN, /* make sure enum is signed */
NETDEV_TX_OK = 0x00, /* driver took care of packet */
NETDEV_TX_BUSY = 0x10, /* driver tx path was busy*/
- NETDEV_TX_LOCKED = 0x20, /* driver tx lock was already taken */
};
typedef enum netdev_tx netdev_tx_t;
@@ -831,7 +830,6 @@ struct tc_to_netdev {
* the queue before that can happen; it's for obsolete devices and weird
* corner cases, but the stack really does a non-trivial amount
* of useless work if you return NETDEV_TX_BUSY.
- * (can also return NETDEV_TX_LOCKED iff NETIF_F_LLTX)
* Required; cannot be NULL.
*
* netdev_features_t (*ndo_fix_features)(struct net_device *dev,
@@ -2737,7 +2735,6 @@ struct softnet_data {
/* stats */
unsigned int processed;
unsigned int time_squeeze;
- unsigned int cpu_collision;
unsigned int received_rps;
#ifdef CONFIG_RPS
struct softnet_data *rps_ipi_list;
diff --git a/net/core/net-procfs.c b/net/core/net-procfs.c
index 2bf8329..14d0934 100644
--- a/net/core/net-procfs.c
+++ b/net/core/net-procfs.c
@@ -162,7 +162,8 @@ static int softnet_seq_show(struct seq_file *seq, void *v)
"%08x %08x %08x %08x %08x %08x %08x %08x %08x %08x %08x\n",
sd->processed, sd->dropped, sd->time_squeeze, 0,
0, 0, 0, 0, /* was fastroute */
- sd->cpu_collision, sd->received_rps, flow_limit_count);
+ 0, /* was cpu_collision */
+ sd->received_rps, flow_limit_count);
return 0;
}
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 20999aa..8604ae2 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -3472,7 +3472,6 @@ xmit_more:
pkt_dev->odevname, ret);
pkt_dev->errors++;
/* fallthru */
- case NETDEV_TX_LOCKED:
case NETDEV_TX_BUSY:
/* Retry it next time */
atomic_dec(&(pkt_dev->skb->users));
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index 80742ed..9c77562 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -108,35 +108,6 @@ static struct sk_buff *dequeue_skb(struct Qdisc *q, bool *validate,
return skb;
}
-static inline int handle_dev_cpu_collision(struct sk_buff *skb,
- struct netdev_queue *dev_queue,
- struct Qdisc *q)
-{
- int ret;
-
- if (unlikely(dev_queue->xmit_lock_owner == smp_processor_id())) {
- /*
- * Same CPU holding the lock. It may be a transient
- * configuration error, when hard_start_xmit() recurses. We
- * detect it by checking xmit owner and drop the packet when
- * deadloop is detected. Return OK to try the next skb.
- */
- kfree_skb_list(skb);
- net_warn_ratelimited("Dead loop on netdevice %s, fix it urgently!\n",
- dev_queue->dev->name);
- ret = qdisc_qlen(q);
- } else {
- /*
- * Another cpu is holding lock, requeue & delay xmits for
- * some time.
- */
- __this_cpu_inc(softnet_data.cpu_collision);
- ret = dev_requeue_skb(skb, q);
- }
-
- return ret;
-}
-
/*
* Transmit possibly several skbs, and handle the return status as
* required. Holding the __QDISC___STATE_RUNNING bit guarantees that
@@ -174,9 +145,6 @@ int sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q,
if (dev_xmit_complete(ret)) {
/* Driver sent out skb successfully or skb was consumed */
ret = qdisc_qlen(q);
- } else if (ret == NETDEV_TX_LOCKED) {
- /* Driver try lock failed */
- ret = handle_dev_cpu_collision(skb, txq, q);
} else {
/* Driver returned NETDEV_TX_BUSY - requeue skb */
if (unlikely(ret != NETDEV_TX_BUSY))
--
2.7.3
^ permalink raw reply related
* [PATCH net-next 5/6] drivers: net: remove NETDEV_TX_LOCKED
From: Florian Westphal @ 2016-04-24 19:38 UTC (permalink / raw)
To: netdev; +Cc: linux-kernel, Florian Westphal, Jon Mason, Andy Gospodarek
In-Reply-To: <1461526694-11367-1-git-send-email-fw@strlen.de>
replace the trylock by a full spin_lock and remove TX_LOCKED return value.
Followup patch will remove TX_LOCKED from the kernel.
Cc: Jon Mason <jdmason@kudzu.us>
Cc: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
drivers/net/ethernet/chelsio/cxgb/sge.c | 3 +--
drivers/net/ethernet/neterion/s2io.c | 9 +--------
drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c | 6 ++----
drivers/net/ethernet/tehuti/tehuti.c | 8 +-------
drivers/net/rionet.c | 6 +-----
5 files changed, 6 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb/sge.c b/drivers/net/ethernet/chelsio/cxgb/sge.c
index 526ea74..86f467a 100644
--- a/drivers/net/ethernet/chelsio/cxgb/sge.c
+++ b/drivers/net/ethernet/chelsio/cxgb/sge.c
@@ -1664,8 +1664,7 @@ static int t1_sge_tx(struct sk_buff *skb, struct adapter *adapter,
struct cmdQ *q = &sge->cmdQ[qid];
unsigned int credits, pidx, genbit, count, use_sched_skb = 0;
- if (!spin_trylock(&q->lock))
- return NETDEV_TX_LOCKED;
+ spin_lock(&q->lock);
reclaim_completed_tx(sge, q);
diff --git a/drivers/net/ethernet/neterion/s2io.c b/drivers/net/ethernet/neterion/s2io.c
index 9ba9758..2874dff 100644
--- a/drivers/net/ethernet/neterion/s2io.c
+++ b/drivers/net/ethernet/neterion/s2io.c
@@ -4021,7 +4021,6 @@ static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev)
unsigned long flags = 0;
u16 vlan_tag = 0;
struct fifo_info *fifo = NULL;
- int do_spin_lock = 1;
int offload_type;
int enable_per_list_interrupt = 0;
struct config_param *config = &sp->config;
@@ -4074,7 +4073,6 @@ static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev)
queue += sp->udp_fifo_idx;
if (skb->len > 1024)
enable_per_list_interrupt = 1;
- do_spin_lock = 0;
}
}
}
@@ -4084,12 +4082,7 @@ static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev)
[skb->priority & (MAX_TX_FIFOS - 1)];
fifo = &mac_control->fifos[queue];
- if (do_spin_lock)
- spin_lock_irqsave(&fifo->tx_lock, flags);
- else {
- if (unlikely(!spin_trylock_irqsave(&fifo->tx_lock, flags)))
- return NETDEV_TX_LOCKED;
- }
+ spin_lock_irqsave(&fifo->tx_lock, flags);
if (sp->config.multiq) {
if (__netif_subqueue_stopped(dev, fifo->fifo_no)) {
diff --git a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c
index 3b98b263b..4475dcc 100644
--- a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c
+++ b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c
@@ -2137,10 +2137,8 @@ static int pch_gbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring;
unsigned long flags;
- if (!spin_trylock_irqsave(&tx_ring->tx_lock, flags)) {
- /* Collision - tell upper layer to requeue */
- return NETDEV_TX_LOCKED;
- }
+ spin_trylock_irqsave(&tx_ring->tx_lock, flags);
+
if (unlikely(!PCH_GBE_DESC_UNUSED(tx_ring))) {
netif_stop_queue(netdev);
spin_unlock_irqrestore(&tx_ring->tx_lock, flags);
diff --git a/drivers/net/ethernet/tehuti/tehuti.c b/drivers/net/ethernet/tehuti/tehuti.c
index 14c9d1b..2524a69 100644
--- a/drivers/net/ethernet/tehuti/tehuti.c
+++ b/drivers/net/ethernet/tehuti/tehuti.c
@@ -1610,7 +1610,6 @@ static inline int bdx_tx_space(struct bdx_priv *priv)
* o NETDEV_TX_BUSY Cannot transmit packet, try later
* Usually a bug, means queue start/stop flow control is broken in
* the driver. Note: the driver must NOT put the skb in its DMA ring.
- * o NETDEV_TX_LOCKED Locking failed, please retry quickly.
*/
static netdev_tx_t bdx_tx_transmit(struct sk_buff *skb,
struct net_device *ndev)
@@ -1630,12 +1629,7 @@ static netdev_tx_t bdx_tx_transmit(struct sk_buff *skb,
ENTER;
local_irq_save(flags);
- if (!spin_trylock(&priv->tx_lock)) {
- local_irq_restore(flags);
- DBG("%s[%s]: TX locked, returning NETDEV_TX_LOCKED\n",
- BDX_DRV_NAME, ndev->name);
- return NETDEV_TX_LOCKED;
- }
+ spin_lock(&priv->tx_lock);
/* build tx descriptor */
BDX_ASSERT(f->m.wptr >= f->m.memsz); /* started with valid wptr */
diff --git a/drivers/net/rionet.c b/drivers/net/rionet.c
index 9cfe6ae..a31f461 100644
--- a/drivers/net/rionet.c
+++ b/drivers/net/rionet.c
@@ -179,11 +179,7 @@ static int rionet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
unsigned long flags;
int add_num = 1;
- local_irq_save(flags);
- if (!spin_trylock(&rnet->tx_lock)) {
- local_irq_restore(flags);
- return NETDEV_TX_LOCKED;
- }
+ spin_lock_irqsave(&rnet->tx_lock, flags);
if (is_multicast_ether_addr(eth->h_dest))
add_num = nets[rnet->mport->id].nact;
--
2.7.3
^ permalink raw reply related
* Re: linux-next: zillions of lockdep whinges in include/net/sock.h:1408
From: Eric Dumazet @ 2016-04-24 19:46 UTC (permalink / raw)
To: Hannes Frederic Sowa; +Cc: David Miller, Valdis.Kletnieks, netdev, linux-kernel
In-Reply-To: <571D14F8.6070306@stressinduktion.org>
On Sun, 2016-04-24 at 20:48 +0200, Hannes Frederic Sowa wrote:
> On 24.04.2016 20:38, David Miller wrote:
> > From: Hannes Frederic Sowa <hannes@stressinduktion.org>
> > Date: Thu, 21 Apr 2016 15:49:37 +0200
> >
> >> On 21.04.2016 15:31, Eric Dumazet wrote:
> >>> On Thu, 2016-04-21 at 05:05 -0400, Valdis.Kletnieks@vt.edu wrote:
> >>>> On Thu, 21 Apr 2016 09:42:12 +0200, Hannes Frederic Sowa said:
> >>>>> Hi,
> >>>>>
> >>>>> On Thu, Apr 21, 2016, at 02:30, Valdis Kletnieks wrote:
> >>>>>> linux-next 20160420 is whining at an incredible rate - in 20 minutes of
> >>>>>> uptime, I piled up some 41,000 hits from all over the place (cleaned up
> >>>>>> to skip the CPU and PID so the list isn't quite so long):
> >>>>>
> >>>>> Thanks for the report. Can you give me some more details:
> >>>>>
> >>>>> Is this an nfs socket? Do you by accident know if this socket went
> >>>>> through xs_reclassify_socket at any point? We do hold the appropriate
> >>>>> locks at that point but I fear that the lockdep reinitialization
> >>>>> confused lockdep.
> >>>>
> >>>> It wasn't an NFS socket, as NFS wasn't even active at the time. I'm reasonably
> >>>> sure that multiple sockets were in play, given that tcp_v6_rcv and
> >>>> udpv6_queue_rcv_skb were both implicated. I strongly suspect that pretty much
> >>>> any IPv6 traffic could do it - the frequency dropped off quite a bit when I
> >>>> closed firefox, which is usually a heavy network hitter on my laptop.
> >>>
> >>>
> >>> Looks like the following patch is needed, can you try it please ?
> >>>
> >>> Thanks !
> >>>
> >>> diff --git a/include/net/sock.h b/include/net/sock.h
> >>> index d997ec13a643..db8301c76d50 100644
> >>> --- a/include/net/sock.h
> >>> +++ b/include/net/sock.h
> >>> @@ -1350,7 +1350,8 @@ static inline bool lockdep_sock_is_held(const struct sock *csk)
> >>> {
> >>> struct sock *sk = (struct sock *)csk;
> >>>
> >>> - return lockdep_is_held(&sk->sk_lock) ||
> >>> + return !debug_locks ||
> >>> + lockdep_is_held(&sk->sk_lock) ||
> >>> lockdep_is_held(&sk->sk_lock.slock);
> >>> }
> >>> #endif
> >>
> >> I would prefer to add debug_locks at the WARN_ON level, like
> >> WARN_ON(debug_locks && !lockdep_sock_is_held(sk)), but I am not sure if
> >> this fixes the initial splat.
> >
> > Can we finish this conversation out and come up with a final patch
> > for this soon?
>
> Eric's patch is worth to apply anyway, but I am not sure if it solves
> the (fundamental) problem. I couldn't reproduce it with the exact next-
> tag provided in the initial mail. All other reports also only happend
> with linux-next and not net-next.
>
> I hope I Valdis provides his config soon and I will continue my analysis
> on this then.
Should be easy to force a lockdep splat and check if the patch solves
the issue.
Issue here is that once lockdep detected a problem (not necessarily in
net/ tree btw), your helper always 'detect' a problem, since lockdep
automatically disables itself.
^ permalink raw reply
* [PATCH net-next V1 01/11] net/mlx5e: Report additional error statistics in get stats ndo
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Provide rtnl_link_stats64 with information regarding physical errors to be
seen in ifconfig and ip tool.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 39 +++++++++++++++++---
1 files changed, 33 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index d485d1e..6270f8d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2074,18 +2074,45 @@ mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats)
{
struct mlx5e_priv *priv = netdev_priv(dev);
struct mlx5e_vport_stats *vstats = &priv->stats.vport;
+ struct mlx5e_pport_stats *pstats = &priv->stats.pport;
stats->rx_packets = vstats->rx_packets;
stats->rx_bytes = vstats->rx_bytes;
stats->tx_packets = vstats->tx_packets;
stats->tx_bytes = vstats->tx_bytes;
- stats->multicast = vstats->rx_multicast_packets +
- vstats->tx_multicast_packets;
- stats->tx_errors = vstats->tx_error_packets;
- stats->rx_errors = vstats->rx_error_packets;
+
+#define PPCNT_GET_802_3_CTR(fld) \
+ (MLX5_GET64(eth_802_3_cntrs_grp_data_layout, \
+ pstats->IEEE_802_3_counters, fld##_high))
+
+#define PPCNT_GET_2863_CTR(fld) \
+ (MLX5_GET64(eth_2863_cntrs_grp_data_layout, \
+ pstats->RFC_2863_counters, fld##_high))
+
+ stats->rx_dropped = priv->stats.qcnt.rx_out_of_buffer;
stats->tx_dropped = vstats->tx_queue_dropped;
- stats->rx_crc_errors = 0;
- stats->rx_length_errors = 0;
+
+ stats->rx_length_errors =
+ PPCNT_GET_802_3_CTR(a_in_range_length_errors) +
+ PPCNT_GET_802_3_CTR(a_out_of_range_length_field) +
+ PPCNT_GET_802_3_CTR(a_frame_too_long_errors);
+ stats->rx_crc_errors =
+ PPCNT_GET_802_3_CTR(a_frame_check_sequence_errors);
+ stats->rx_frame_errors =
+ PPCNT_GET_802_3_CTR(a_alignment_errors);
+ stats->tx_aborted_errors =
+ PPCNT_GET_2863_CTR(if_out_discards);
+ stats->tx_carrier_errors =
+ PPCNT_GET_802_3_CTR(a_symbol_error_during_carrier);
+ stats->rx_errors = stats->rx_length_errors + stats->rx_crc_errors +
+ stats->rx_frame_errors;
+ stats->tx_errors = stats->tx_aborted_errors + stats->tx_carrier_errors;
+
+ /* vport multicast also counts packets that are dropped due to steering
+ * or rx out of buffer
+ */
+ stats->multicast = vstats->rx_multicast_packets;
+
return stats;
}
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 00/11] Mellanox 100G extending mlx5 ethtool support
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Saeed Mahameed
Hi Dave,
Changes from V0:
- Dropped: net/mlx5e: Disable link up on INIT HCA command
Due to Ido's and Or's requests we will submit this patch to net and will need it for -stable.
- Rebased to: 11afbff86168 ("Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next")
This series is centralized around extending and improving mlx5 ethernet driver ethtool
support. We've done some code refactoring for ethtool statistics reporting, making it
more scalable and robust, now each reported ethtool counter belongs to a group and has
its own descriptor within that group, the descriptor holds the counter name and offset
in memory in that group memory block.
Added new counters:
- Reporting more error and drop counter in ifconig/ip tool.
- Per priority pause and traffic counter in ethtool.
- link down events counter in ethtool.
Set features handling was also refactored a little bit to be more resilient and generic,
now setting more than one feature will not stop on the first failed one, but instead
it will try to continue setting others. We made it generic to make it simpler for adding
more features support, it is now done easily by only introducing a handler function of
the new supported netdev feature, and let the generic handler do the job.
New netdev features and ethtool support:
- Netdev feature RXALL, set on/off FCS check offload.
- Netdev feature HW_VLAN_CTAG_RX, set on/off rx-vlan stripping offload.
- Ethtool interface identify.
- Ethtool dump module EEPROM.
Thanks,
Saeed
Eran Ben Elisha (1):
net/mlx5e: Add support for RXALL netdev feature
Gal Pressman (9):
net/mlx5e: Report additional error statistics in get stats ndo
net/mlx5e: Statistics handling refactoring
net/mlx5e: Rename VPort counters
net/mlx5e: Add per priority group to PPort counters
net/mlx5e: Add link down events counter
net/mlx5e: Improve set features ndo resiliency
net/mlx5e: Add ethtool support for interface identify (LED blinking)
net/mlx5e: Add ethtool support for dump module EEPROM
net/mlx5e: Add ethtool support for rxvlan-offload (vlan stripping)
Saeed Mahameed (1):
net/mlx5e: Fix checksum handling for non-stripped vlan packets
drivers/net/ethernet/mellanox/mlx5/core/en.h | 243 +----------
drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c | 6 +
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 284 ++++++++++--
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 455 ++++++++++++--------
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 20 +-
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 359 +++++++++++++++
drivers/net/ethernet/mellanox/mlx5/core/port.c | 138 ++++++
include/linux/mlx5/device.h | 23 +-
include/linux/mlx5/driver.h | 9 +-
include/linux/mlx5/port.h | 25 ++
10 files changed, 1092 insertions(+), 470 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
^ permalink raw reply
* [PATCH net-next V1 03/11] net/mlx5e: Rename VPort counters
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
VPort and software counters names are confusing and may be unclear, all
VPort counters now have a prefix of rx/tx_vport_*.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 35 +++++++++++---------
1 files changed, 19 insertions(+), 16 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index 116320d..4f3a08d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -115,32 +115,35 @@ struct mlx5e_vport_stats {
};
static const struct counter_desc vport_stats_desc[] = {
- { "rx_error_packets", VPORT_COUNTER_OFF(received_errors.packets) },
- { "rx_error_bytes", VPORT_COUNTER_OFF(received_errors.octets) },
- { "tx_error_packets", VPORT_COUNTER_OFF(transmit_errors.packets) },
- { "tx_error_bytes", VPORT_COUNTER_OFF(transmit_errors.octets) },
- { "rx_unicast_packets",
+ { "rx_vport_error_packets",
+ VPORT_COUNTER_OFF(received_errors.packets) },
+ { "rx_vport_error_bytes", VPORT_COUNTER_OFF(received_errors.octets) },
+ { "tx_vport_error_packets",
+ VPORT_COUNTER_OFF(transmit_errors.packets) },
+ { "tx_vport_error_bytes", VPORT_COUNTER_OFF(transmit_errors.octets) },
+ { "rx_vport_unicast_packets",
VPORT_COUNTER_OFF(received_eth_unicast.packets) },
- { "rx_unicast_bytes", VPORT_COUNTER_OFF(received_eth_unicast.octets) },
- { "tx_unicast_packets",
+ { "rx_vport_unicast_bytes",
+ VPORT_COUNTER_OFF(received_eth_unicast.octets) },
+ { "tx_vport_unicast_packets",
VPORT_COUNTER_OFF(transmitted_eth_unicast.packets) },
- { "tx_unicast_bytes",
+ { "tx_vport_unicast_bytes",
VPORT_COUNTER_OFF(transmitted_eth_unicast.octets) },
- { "rx_multicast_packets",
+ { "rx_vport_multicast_packets",
VPORT_COUNTER_OFF(received_eth_multicast.packets) },
- { "rx_multicast_bytes",
+ { "rx_vport_multicast_bytes",
VPORT_COUNTER_OFF(received_eth_multicast.octets) },
- { "tx_multicast_packets",
+ { "tx_vport_multicast_packets",
VPORT_COUNTER_OFF(transmitted_eth_multicast.packets) },
- { "tx_multicast_bytes",
+ { "tx_vport_multicast_bytes",
VPORT_COUNTER_OFF(transmitted_eth_multicast.octets) },
- { "rx_broadcast_packets",
+ { "rx_vport_broadcast_packets",
VPORT_COUNTER_OFF(received_eth_broadcast.packets) },
- { "rx_broadcast_bytes",
+ { "rx_vport_broadcast_bytes",
VPORT_COUNTER_OFF(received_eth_broadcast.octets) },
- { "tx_broadcast_packets",
+ { "tx_vport_broadcast_packets",
VPORT_COUNTER_OFF(transmitted_eth_broadcast.packets) },
- { "tx_broadcast_bytes",
+ { "tx_vport_broadcast_bytes",
VPORT_COUNTER_OFF(transmitted_eth_broadcast.octets) },
};
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 05/11] net/mlx5e: Add link down events counter
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Expose link_down_events counter through ethtool -S.
This counter is read from PPort statistics, then proccessed and stored as
a special handling software counter.
This counter is stored along software counters since it is the only PPort
counter that it's size is not 64 bits.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 10 +++++++++-
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 5 +++++
include/linux/mlx5/device.h | 1 +
3 files changed, 15 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index ef66ba6..61e261c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -135,6 +135,10 @@ static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
s->tx_csum_offload = s->tx_packets - tx_offload_none - s->tx_csum_inner;
s->rx_csum_good = s->rx_packets - s->rx_csum_none -
s->rx_csum_sw;
+
+ s->link_down_events = MLX5_GET(ppcnt_reg,
+ priv->stats.pport.phy_counters,
+ counter_set.phys_layer_cntrs.link_down_events);
}
static void mlx5e_update_vport_counters(struct mlx5e_priv *priv)
@@ -183,6 +187,10 @@ static void mlx5e_update_pport_counters(struct mlx5e_priv *priv)
MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP);
mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
+ out = pstats->phy_counters;
+ MLX5_SET(ppcnt_reg, in, grp, MLX5_PHYSICAL_LAYER_COUNTERS_GROUP);
+ mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
+
MLX5_SET(ppcnt_reg, in, grp, MLX5_PER_PRIORITY_COUNTERS_GROUP);
for (prio = 0; prio < NUM_PPORT_PRIO; prio++) {
out = pstats->per_prio_counters[prio];
@@ -208,10 +216,10 @@ static void mlx5e_update_q_counter(struct mlx5e_priv *priv)
void mlx5e_update_stats(struct mlx5e_priv *priv)
{
- mlx5e_update_sw_counters(priv);
mlx5e_update_q_counter(priv);
mlx5e_update_vport_counters(priv);
mlx5e_update_pport_counters(priv);
+ mlx5e_update_sw_counters(priv);
}
static void mlx5e_update_stats_work(struct work_struct *work)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index de27eea..7cd8cb4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -71,6 +71,9 @@ struct mlx5e_sw_stats {
u64 rx_mpwqe_filler;
u64 rx_mpwqe_frag;
u64 rx_buff_alloc_err;
+
+ /* Special handling counters */
+ u64 link_down_events;
};
static const struct counter_desc sw_stats_desc[] = {
@@ -96,6 +99,7 @@ static const struct counter_desc sw_stats_desc[] = {
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_mpwqe_filler) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_mpwqe_frag) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_buff_alloc_err) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, link_down_events) },
};
struct mlx5e_qcounter_stats {
@@ -178,6 +182,7 @@ struct mlx5e_pport_stats {
__be64 RFC_2863_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
__be64 RFC_2819_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
__be64 per_prio_counters[NUM_PPORT_PRIO][MLX5_ST_SZ_QW(ppcnt_reg)];
+ __be64 phy_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
};
static const struct counter_desc pport_802_3_stats_desc[] = {
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 8be44ca..942bcca 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -1369,6 +1369,7 @@ enum {
MLX5_ETHERNET_EXTENDED_COUNTERS_GROUP = 0x5,
MLX5_PER_PRIORITY_COUNTERS_GROUP = 0x10,
MLX5_PER_TRAFFIC_CLASS_COUNTERS_GROUP = 0x11,
+ MLX5_PHYSICAL_LAYER_COUNTERS_GROUP = 0x12,
MLX5_INFINIBAND_PORT_COUNTERS_GROUP = 0x20,
};
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 07/11] net/mlx5e: Add support for RXALL netdev feature
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Eran Ben Elisha <eranbe@mellanox.com>
Introduce new access register named Ports Check Mask Register (PCMR) to
control all HW checks on port. With this register, the driver can
enable/disable Hardware FCS validation.
When RXALL is enabled/disabled using ndo_set_features, enable/disable
fcs check at HW.
User can change HW configuration using rx-all flag at ethtool.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 20 ++++++++
drivers/net/ethernet/mellanox/mlx5/core/port.c | 49 +++++++++++++++++++++
include/linux/mlx5/driver.h | 1 +
include/linux/mlx5/port.h | 4 ++
4 files changed, 74 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index d82bc6b..ad0cb4a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2139,6 +2139,14 @@ static int set_feature_tc_num_filters(struct net_device *netdev, bool enable)
return 0;
}
+static int set_feature_rx_all(struct net_device *netdev, bool enable)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+ struct mlx5_core_dev *mdev = priv->mdev;
+
+ return mlx5_set_port_fcs(mdev, !enable);
+}
+
static int mlx5e_handle_feature(struct net_device *netdev,
netdev_features_t wanted_features,
netdev_features_t feature,
@@ -2174,6 +2182,8 @@ static int mlx5e_set_features(struct net_device *netdev,
set_feature_vlan_filter);
err |= mlx5e_handle_feature(netdev, features, NETIF_F_HW_TC,
set_feature_tc_num_filters);
+ err |= mlx5e_handle_feature(netdev, features, NETIF_F_RXALL,
+ set_feature_rx_all);
return err ? -EINVAL : 0;
}
@@ -2564,6 +2574,8 @@ static void mlx5e_build_netdev(struct net_device *netdev)
{
struct mlx5e_priv *priv = netdev_priv(netdev);
struct mlx5_core_dev *mdev = priv->mdev;
+ bool fcs_supported;
+ bool fcs_enabled;
SET_NETDEV_DEV(netdev, &mdev->pdev->dev);
@@ -2607,10 +2619,18 @@ static void mlx5e_build_netdev(struct net_device *netdev)
netdev->hw_enc_features |= NETIF_F_GSO_UDP_TUNNEL;
}
+ mlx5_query_port_fcs(mdev, &fcs_supported, &fcs_enabled);
+
+ if (fcs_supported)
+ netdev->hw_features |= NETIF_F_RXALL;
+
netdev->features = netdev->hw_features;
if (!priv->params.lro_en)
netdev->features &= ~NETIF_F_LRO;
+ if (fcs_enabled)
+ netdev->features &= ~NETIF_F_RXALL;
+
#define FT_CAP(f) MLX5_CAP_FLOWTABLE(mdev, flow_table_properties_nic_receive.f)
if (FT_CAP(flow_modify_en) &&
FT_CAP(modify_root) &&
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c
index ae378c5..c37740f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c
@@ -607,3 +607,52 @@ int mlx5_query_port_wol(struct mlx5_core_dev *mdev, u8 *wol_mode)
return err;
}
EXPORT_SYMBOL_GPL(mlx5_query_port_wol);
+
+static int mlx5_query_ports_check(struct mlx5_core_dev *mdev, u32 *out,
+ int outlen)
+{
+ u32 in[MLX5_ST_SZ_DW(pcmr_reg)];
+
+ memset(in, 0, sizeof(in));
+ MLX5_SET(pcmr_reg, in, local_port, 1);
+
+ return mlx5_core_access_reg(mdev, in, sizeof(in), out,
+ outlen, MLX5_REG_PCMR, 0, 0);
+}
+
+static int mlx5_set_ports_check(struct mlx5_core_dev *mdev, u32 *in, int inlen)
+{
+ u32 out[MLX5_ST_SZ_DW(pcmr_reg)];
+
+ return mlx5_core_access_reg(mdev, in, inlen, out,
+ sizeof(out), MLX5_REG_PCMR, 0, 1);
+}
+
+int mlx5_set_port_fcs(struct mlx5_core_dev *mdev, u8 enable)
+{
+ u32 in[MLX5_ST_SZ_DW(pcmr_reg)];
+
+ memset(in, 0, sizeof(in));
+ MLX5_SET(pcmr_reg, in, local_port, 1);
+ MLX5_SET(pcmr_reg, in, fcs_chk, enable);
+
+ return mlx5_set_ports_check(mdev, in, sizeof(in));
+}
+
+void mlx5_query_port_fcs(struct mlx5_core_dev *mdev, bool *supported,
+ bool *enabled)
+{
+ u32 out[MLX5_ST_SZ_DW(pcmr_reg)];
+ /* Default values for FW which do not support MLX5_REG_PCMR */
+ *supported = false;
+ *enabled = true;
+
+ if (!MLX5_CAP_GEN(mdev, ports_check))
+ return;
+
+ if (mlx5_query_ports_check(mdev, out, sizeof(out)))
+ return;
+
+ *supported = !!(MLX5_GET(pcmr_reg, out, fcs_cap));
+ *enabled = !!(MLX5_GET(pcmr_reg, out, fcs_chk));
+}
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index dcd5ac8..497a4db 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -112,6 +112,7 @@ enum {
MLX5_REG_PMPE = 0x5010,
MLX5_REG_PELC = 0x500e,
MLX5_REG_PVLC = 0x500f,
+ MLX5_REG_PCMR = 0x5041,
MLX5_REG_PMLP = 0, /* TBD */
MLX5_REG_NODE_DESC = 0x6001,
MLX5_REG_HOST_ENDIANNESS = 0x7004,
diff --git a/include/linux/mlx5/port.h b/include/linux/mlx5/port.h
index a1d145a..577e953 100644
--- a/include/linux/mlx5/port.h
+++ b/include/linux/mlx5/port.h
@@ -84,4 +84,8 @@ int mlx5_query_port_ets_rate_limit(struct mlx5_core_dev *mdev,
int mlx5_set_port_wol(struct mlx5_core_dev *mdev, u8 wol_mode);
int mlx5_query_port_wol(struct mlx5_core_dev *mdev, u8 *wol_mode);
+int mlx5_set_port_fcs(struct mlx5_core_dev *mdev, u8 enable);
+void mlx5_query_port_fcs(struct mlx5_core_dev *mdev, bool *supported,
+ bool *enabled);
+
#endif /* __MLX5_PORT_H__ */
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 06/11] net/mlx5e: Improve set features ndo resiliency
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
In current mlx5e ndo_set_features implementation, setting some features
can success while others can fail. Today, we return one error code which
doesn't reflect the current features status of the netdev at the end of
the ndo callback.
Set netdev->features with features which were successfully set in order
to keep the current status in case of failure. For this purpose, define
new Macro to set/unset specific feature in netdev->features.
This patch introduces a mechanism that uses feature handlers for each
feature.
Set features will call a generic handler, which will then call a specific
handler in his turn and update netdev->features according to it's return
value. Each specific handler is responsible to perform driver specific
actions, and updating params if needed.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 110 +++++++++++++++-----
1 files changed, 82 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 61e261c..d82bc6b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2078,50 +2078,104 @@ static int mlx5e_set_mac(struct net_device *netdev, void *addr)
return 0;
}
-static int mlx5e_set_features(struct net_device *netdev,
- netdev_features_t features)
+#define MLX5E_SET_FEATURE(netdev, feature, enable) \
+ do { \
+ if (enable) \
+ netdev->features |= feature; \
+ else \
+ netdev->features &= ~feature; \
+ } while (0)
+
+typedef int (*mlx5e_feature_handler)(struct net_device *netdev, bool enable);
+
+static int set_feature_lro(struct net_device *netdev, bool enable)
{
struct mlx5e_priv *priv = netdev_priv(netdev);
- int err = 0;
- netdev_features_t changes = features ^ netdev->features;
+ bool was_opened = test_bit(MLX5E_STATE_OPENED, &priv->state);
+ int err;
mutex_lock(&priv->state_lock);
- if (changes & NETIF_F_LRO) {
- bool was_opened = test_bit(MLX5E_STATE_OPENED, &priv->state);
-
- if (was_opened && (priv->params.rq_wq_type ==
- MLX5_WQ_TYPE_LINKED_LIST))
- mlx5e_close_locked(priv->netdev);
-
- priv->params.lro_en = !!(features & NETIF_F_LRO);
- err = mlx5e_modify_tirs_lro(priv);
- if (err)
- mlx5_core_warn(priv->mdev, "lro modify failed, %d\n",
- err);
+ if (was_opened && (priv->params.rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST))
+ mlx5e_close_locked(priv->netdev);
- if (was_opened && (priv->params.rq_wq_type ==
- MLX5_WQ_TYPE_LINKED_LIST))
- err = mlx5e_open_locked(priv->netdev);
+ priv->params.lro_en = enable;
+ err = mlx5e_modify_tirs_lro(priv);
+ if (err) {
+ netdev_err(netdev, "lro modify failed, %d\n", err);
+ priv->params.lro_en = !enable;
}
+ if (was_opened && (priv->params.rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST))
+ mlx5e_open_locked(priv->netdev);
+
mutex_unlock(&priv->state_lock);
- if (changes & NETIF_F_HW_VLAN_CTAG_FILTER) {
- if (features & NETIF_F_HW_VLAN_CTAG_FILTER)
- mlx5e_enable_vlan_filter(priv);
- else
- mlx5e_disable_vlan_filter(priv);
- }
+ return err;
+}
+
+static int set_feature_vlan_filter(struct net_device *netdev, bool enable)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+
+ if (enable)
+ mlx5e_enable_vlan_filter(priv);
+ else
+ mlx5e_disable_vlan_filter(priv);
+
+ return 0;
+}
+
+static int set_feature_tc_num_filters(struct net_device *netdev, bool enable)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
- if ((changes & NETIF_F_HW_TC) && !(features & NETIF_F_HW_TC) &&
- mlx5e_tc_num_filters(priv)) {
+ if (!enable && mlx5e_tc_num_filters(priv)) {
netdev_err(netdev,
"Active offloaded tc filters, can't turn hw_tc_offload off\n");
return -EINVAL;
}
- return err;
+ return 0;
+}
+
+static int mlx5e_handle_feature(struct net_device *netdev,
+ netdev_features_t wanted_features,
+ netdev_features_t feature,
+ mlx5e_feature_handler feature_handler)
+{
+ netdev_features_t changes = wanted_features ^ netdev->features;
+ bool enable = !!(wanted_features & feature);
+ int err;
+
+ if (!(changes & feature))
+ return 0;
+
+ err = feature_handler(netdev, enable);
+ if (err) {
+ netdev_err(netdev, "%s feature 0x%llx failed err %d\n",
+ enable ? "Enable" : "Disable", feature, err);
+ return err;
+ }
+
+ MLX5E_SET_FEATURE(netdev, feature, enable);
+ return 0;
+}
+
+static int mlx5e_set_features(struct net_device *netdev,
+ netdev_features_t features)
+{
+ int err;
+
+ err = mlx5e_handle_feature(netdev, features, NETIF_F_LRO,
+ set_feature_lro);
+ err |= mlx5e_handle_feature(netdev, features,
+ NETIF_F_HW_VLAN_CTAG_FILTER,
+ set_feature_vlan_filter);
+ err |= mlx5e_handle_feature(netdev, features, NETIF_F_HW_TC,
+ set_feature_tc_num_filters);
+
+ return err ? -EINVAL : 0;
}
static int mlx5e_change_mtu(struct net_device *netdev, int new_mtu)
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 09/11] net/mlx5e: Add ethtool support for dump module EEPROM
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Add query MCIA, PMLP registers infrastructure and commands.
Add ethtool support for get_module_info() and get_module_eeprom()
callbacks.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 80 ++++++++++++++++++++
drivers/net/ethernet/mellanox/mlx5/core/port.c | 76 +++++++++++++++++++
include/linux/mlx5/driver.h | 3 +-
include/linux/mlx5/port.h | 15 ++++
4 files changed, 173 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index a2c444e..0518c86 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -1159,6 +1159,84 @@ static int mlx5e_set_phys_id(struct net_device *dev,
return mlx5_set_port_beacon(mdev, beacon_duration);
}
+static int mlx5e_get_module_info(struct net_device *netdev,
+ struct ethtool_modinfo *modinfo)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+ struct mlx5_core_dev *dev = priv->mdev;
+ int size_read = 0;
+ u8 data[4];
+
+ size_read = mlx5_query_module_eeprom(dev, 0, 2, data);
+ if (size_read < 2)
+ return -EIO;
+
+ /* data[0] = identifier byte */
+ switch (data[0]) {
+ case MLX5_MODULE_ID_QSFP:
+ modinfo->type = ETH_MODULE_SFF_8436;
+ modinfo->eeprom_len = ETH_MODULE_SFF_8436_LEN;
+ break;
+ case MLX5_MODULE_ID_QSFP_PLUS:
+ case MLX5_MODULE_ID_QSFP28:
+ /* data[1] = revision id */
+ if (data[0] == MLX5_MODULE_ID_QSFP28 || data[1] >= 0x3) {
+ modinfo->type = ETH_MODULE_SFF_8636;
+ modinfo->eeprom_len = ETH_MODULE_SFF_8636_LEN;
+ } else {
+ modinfo->type = ETH_MODULE_SFF_8436;
+ modinfo->eeprom_len = ETH_MODULE_SFF_8436_LEN;
+ }
+ break;
+ case MLX5_MODULE_ID_SFP:
+ modinfo->type = ETH_MODULE_SFF_8472;
+ modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN;
+ break;
+ default:
+ netdev_err(priv->netdev, "%s: cable type not recognized:0x%x\n",
+ __func__, data[0]);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int mlx5e_get_module_eeprom(struct net_device *netdev,
+ struct ethtool_eeprom *ee,
+ u8 *data)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+ struct mlx5_core_dev *mdev = priv->mdev;
+ int offset = ee->offset;
+ int size_read;
+ int i = 0;
+
+ if (!ee->len)
+ return -EINVAL;
+
+ memset(data, 0, ee->len);
+
+ while (i < ee->len) {
+ size_read = mlx5_query_module_eeprom(mdev, offset, ee->len - i,
+ data + i);
+
+ if (!size_read)
+ /* Done reading */
+ return 0;
+
+ if (size_read < 0) {
+ netdev_err(priv->netdev, "%s: mlx5_query_eeprom failed:0x%x\n",
+ __func__, size_read);
+ return 0;
+ }
+
+ i += size_read;
+ offset += size_read;
+ }
+
+ return 0;
+}
+
const struct ethtool_ops mlx5e_ethtool_ops = {
.get_drvinfo = mlx5e_get_drvinfo,
.get_link = ethtool_op_get_link,
@@ -1186,4 +1264,6 @@ const struct ethtool_ops mlx5e_ethtool_ops = {
.set_phys_id = mlx5e_set_phys_id,
.get_wol = mlx5e_get_wol,
.set_wol = mlx5e_set_wol,
+ .get_module_info = mlx5e_get_module_info,
+ .get_module_eeprom = mlx5e_get_module_eeprom,
};
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c
index 446549f..4cb2a44 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c
@@ -310,6 +310,82 @@ void mlx5_query_port_oper_mtu(struct mlx5_core_dev *dev, int *oper_mtu,
}
EXPORT_SYMBOL_GPL(mlx5_query_port_oper_mtu);
+static int mlx5_query_module_num(struct mlx5_core_dev *dev, int *module_num)
+{
+ u32 out[MLX5_ST_SZ_DW(pmlp_reg)];
+ u32 in[MLX5_ST_SZ_DW(pmlp_reg)];
+ int module_mapping;
+ int err;
+
+ memset(in, 0, sizeof(in));
+
+ MLX5_SET(pmlp_reg, in, local_port, 1);
+
+ err = mlx5_core_access_reg(dev, in, sizeof(in), out, sizeof(out),
+ MLX5_REG_PMLP, 0, 0);
+ if (err)
+ return err;
+
+ module_mapping = MLX5_GET(pmlp_reg, out, lane0_module_mapping);
+ *module_num = module_mapping & MLX5_EEPROM_IDENTIFIER_BYTE_MASK;
+
+ return 0;
+}
+
+int mlx5_query_module_eeprom(struct mlx5_core_dev *dev,
+ u16 offset, u16 size, u8 *data)
+{
+ u32 out[MLX5_ST_SZ_DW(mcia_reg)];
+ u32 in[MLX5_ST_SZ_DW(mcia_reg)];
+ int module_num;
+ u16 i2c_addr;
+ int status;
+ int err;
+ void *ptr = MLX5_ADDR_OF(mcia_reg, out, dword_0);
+
+ err = mlx5_query_module_num(dev, &module_num);
+ if (err)
+ return err;
+
+ memset(in, 0, sizeof(in));
+ size = min_t(int, size, MLX5_EEPROM_MAX_BYTES);
+
+ if (offset < MLX5_EEPROM_PAGE_LENGTH &&
+ offset + size > MLX5_EEPROM_PAGE_LENGTH)
+ /* Cross pages read, read until offset 256 in low page */
+ size -= offset + size - MLX5_EEPROM_PAGE_LENGTH;
+
+ i2c_addr = MLX5_I2C_ADDR_LOW;
+ if (offset >= MLX5_EEPROM_PAGE_LENGTH) {
+ i2c_addr = MLX5_I2C_ADDR_HIGH;
+ offset -= MLX5_EEPROM_PAGE_LENGTH;
+ }
+
+ MLX5_SET(mcia_reg, in, l, 0);
+ MLX5_SET(mcia_reg, in, module, module_num);
+ MLX5_SET(mcia_reg, in, i2c_device_address, i2c_addr);
+ MLX5_SET(mcia_reg, in, page_number, 0);
+ MLX5_SET(mcia_reg, in, device_address, offset);
+ MLX5_SET(mcia_reg, in, size, size);
+
+ err = mlx5_core_access_reg(dev, in, sizeof(in), out,
+ sizeof(out), MLX5_REG_MCIA, 0, 0);
+ if (err)
+ return err;
+
+ status = MLX5_GET(mcia_reg, out, status);
+ if (status) {
+ mlx5_core_err(dev, "query_mcia_reg failed: status: 0x%x\n",
+ status);
+ return -EIO;
+ }
+
+ memcpy(data, ptr, size);
+
+ return size;
+}
+EXPORT_SYMBOL_GPL(mlx5_query_module_eeprom);
+
static int mlx5_query_port_pvlc(struct mlx5_core_dev *dev, u32 *pvlc,
int pvlc_size, u8 local_port)
{
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 2e8758d..1a17067 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -113,9 +113,10 @@ enum {
MLX5_REG_PELC = 0x500e,
MLX5_REG_PVLC = 0x500f,
MLX5_REG_PCMR = 0x5041,
- MLX5_REG_PMLP = 0, /* TBD */
+ MLX5_REG_PMLP = 0x5002,
MLX5_REG_NODE_DESC = 0x6001,
MLX5_REG_HOST_ENDIANNESS = 0x7004,
+ MLX5_REG_MCIA = 0x9014,
MLX5_REG_MLCR = 0x902b,
};
diff --git a/include/linux/mlx5/port.h b/include/linux/mlx5/port.h
index a364ab1..7391eb8 100644
--- a/include/linux/mlx5/port.h
+++ b/include/linux/mlx5/port.h
@@ -40,6 +40,19 @@ enum mlx5_beacon_duration {
MLX5_BEACON_DURATION_INF = 0xffff,
};
+enum mlx5_module_id {
+ MLX5_MODULE_ID_SFP = 0x3,
+ MLX5_MODULE_ID_QSFP = 0xC,
+ MLX5_MODULE_ID_QSFP_PLUS = 0xD,
+ MLX5_MODULE_ID_QSFP28 = 0x11,
+};
+
+#define MLX5_EEPROM_MAX_BYTES 32
+#define MLX5_EEPROM_IDENTIFIER_BYTE_MASK 0x000000ff
+#define MLX5_I2C_ADDR_LOW 0x50
+#define MLX5_I2C_ADDR_HIGH 0x51
+#define MLX5_EEPROM_PAGE_LENGTH 256
+
int mlx5_set_port_caps(struct mlx5_core_dev *dev, u8 port_num, u32 caps);
int mlx5_query_port_ptys(struct mlx5_core_dev *dev, u32 *ptys,
int ptys_size, int proto_mask, u8 local_port);
@@ -93,5 +106,7 @@ int mlx5_query_port_wol(struct mlx5_core_dev *mdev, u8 *wol_mode);
int mlx5_set_port_fcs(struct mlx5_core_dev *mdev, u8 enable);
void mlx5_query_port_fcs(struct mlx5_core_dev *mdev, bool *supported,
bool *enabled);
+int mlx5_query_module_eeprom(struct mlx5_core_dev *dev,
+ u16 offset, u16 size, u8 *data);
#endif /* __MLX5_PORT_H__ */
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 02/11] net/mlx5e: Statistics handling refactoring
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Redesign ethtool statistics handling and reporting in the driver:
1. Move counters to a separate file (en_stats.h).
2. Remove unnecessary dependencies between stats and strings.
3. Use counter descriptors which hold a name and offset for each counter,
and will be used to decide which counters will be exposed.
For example when adding a new software counter to ethtool, instead of:
1. Add to stats struct.
2. Add to strings struct in the same order.
3. Change macro defining number of software counters.
The only thing needed is to link the new counter to a counter descriptor.
VPort counters are a set of hardware traffic counters created automatically
for each virtual port opened.
PPort counters are a set of counters describing per physical port
performance statistics.
These counters are gathered from hardware register and divided to groups
according to different protocols.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 240 +---------------
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 132 ++++++---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 232 +++++----------
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 318 ++++++++++++++++++++
include/linux/mlx5/device.h | 1 +
5 files changed, 483 insertions(+), 440 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 6e24e82..e903eff 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -46,6 +46,7 @@
#include <linux/rhashtable.h>
#include "wq.h"
#include "mlx5_core.h"
+#include "en_stats.h"
#define MLX5E_MAX_NUM_TC 8
@@ -148,245 +149,6 @@ struct mlx5e_umr_wqe {
#define MLX5E_MIN_BW_ALLOC 1 /* Min percentage of BW allocation */
#endif
-static const char vport_strings[][ETH_GSTRING_LEN] = {
- /* vport statistics */
- "rx_packets",
- "rx_bytes",
- "tx_packets",
- "tx_bytes",
- "rx_error_packets",
- "rx_error_bytes",
- "tx_error_packets",
- "tx_error_bytes",
- "rx_unicast_packets",
- "rx_unicast_bytes",
- "tx_unicast_packets",
- "tx_unicast_bytes",
- "rx_multicast_packets",
- "rx_multicast_bytes",
- "tx_multicast_packets",
- "tx_multicast_bytes",
- "rx_broadcast_packets",
- "rx_broadcast_bytes",
- "tx_broadcast_packets",
- "tx_broadcast_bytes",
-
- /* SW counters */
- "tso_packets",
- "tso_bytes",
- "tso_inner_packets",
- "tso_inner_bytes",
- "lro_packets",
- "lro_bytes",
- "rx_csum_good",
- "rx_csum_none",
- "rx_csum_sw",
- "tx_csum_offload",
- "tx_csum_inner",
- "tx_queue_stopped",
- "tx_queue_wake",
- "tx_queue_dropped",
- "rx_wqe_err",
- "rx_mpwqe_filler",
- "rx_mpwqe_frag",
- "rx_buff_alloc_err",
-};
-
-struct mlx5e_vport_stats {
- /* HW counters */
- u64 rx_packets;
- u64 rx_bytes;
- u64 tx_packets;
- u64 tx_bytes;
- u64 rx_error_packets;
- u64 rx_error_bytes;
- u64 tx_error_packets;
- u64 tx_error_bytes;
- u64 rx_unicast_packets;
- u64 rx_unicast_bytes;
- u64 tx_unicast_packets;
- u64 tx_unicast_bytes;
- u64 rx_multicast_packets;
- u64 rx_multicast_bytes;
- u64 tx_multicast_packets;
- u64 tx_multicast_bytes;
- u64 rx_broadcast_packets;
- u64 rx_broadcast_bytes;
- u64 tx_broadcast_packets;
- u64 tx_broadcast_bytes;
-
- /* SW counters */
- u64 tso_packets;
- u64 tso_bytes;
- u64 tso_inner_packets;
- u64 tso_inner_bytes;
- u64 lro_packets;
- u64 lro_bytes;
- u64 rx_csum_good;
- u64 rx_csum_none;
- u64 rx_csum_sw;
- u64 tx_csum_offload;
- u64 tx_csum_inner;
- u64 tx_queue_stopped;
- u64 tx_queue_wake;
- u64 tx_queue_dropped;
- u64 rx_wqe_err;
- u64 rx_mpwqe_filler;
- u64 rx_mpwqe_frag;
- u64 rx_buff_alloc_err;
-
-#define NUM_VPORT_COUNTERS 38
-};
-
-static const char pport_strings[][ETH_GSTRING_LEN] = {
- /* IEEE802.3 counters */
- "frames_tx",
- "frames_rx",
- "check_seq_err",
- "alignment_err",
- "octets_tx",
- "octets_received",
- "multicast_xmitted",
- "broadcast_xmitted",
- "multicast_rx",
- "broadcast_rx",
- "in_range_len_errors",
- "out_of_range_len",
- "too_long_errors",
- "symbol_err",
- "mac_control_tx",
- "mac_control_rx",
- "unsupported_op_rx",
- "pause_ctrl_rx",
- "pause_ctrl_tx",
-
- /* RFC2863 counters */
- "in_octets",
- "in_ucast_pkts",
- "in_discards",
- "in_errors",
- "in_unknown_protos",
- "out_octets",
- "out_ucast_pkts",
- "out_discards",
- "out_errors",
- "in_multicast_pkts",
- "in_broadcast_pkts",
- "out_multicast_pkts",
- "out_broadcast_pkts",
-
- /* RFC2819 counters */
- "drop_events",
- "octets",
- "pkts",
- "broadcast_pkts",
- "multicast_pkts",
- "crc_align_errors",
- "undersize_pkts",
- "oversize_pkts",
- "fragments",
- "jabbers",
- "collisions",
- "p64octets",
- "p65to127octets",
- "p128to255octets",
- "p256to511octets",
- "p512to1023octets",
- "p1024to1518octets",
- "p1519to2047octets",
- "p2048to4095octets",
- "p4096to8191octets",
- "p8192to10239octets",
-};
-
-#define NUM_IEEE_802_3_COUNTERS 19
-#define NUM_RFC_2863_COUNTERS 13
-#define NUM_RFC_2819_COUNTERS 21
-#define NUM_PPORT_COUNTERS (NUM_IEEE_802_3_COUNTERS + \
- NUM_RFC_2863_COUNTERS + \
- NUM_RFC_2819_COUNTERS)
-
-struct mlx5e_pport_stats {
- __be64 IEEE_802_3_counters[NUM_IEEE_802_3_COUNTERS];
- __be64 RFC_2863_counters[NUM_RFC_2863_COUNTERS];
- __be64 RFC_2819_counters[NUM_RFC_2819_COUNTERS];
-};
-
-static const char qcounter_stats_strings[][ETH_GSTRING_LEN] = {
- "rx_out_of_buffer",
-};
-
-struct mlx5e_qcounter_stats {
- u32 rx_out_of_buffer;
-#define NUM_Q_COUNTERS 1
-};
-
-static const char rq_stats_strings[][ETH_GSTRING_LEN] = {
- "packets",
- "bytes",
- "csum_none",
- "csum_sw",
- "lro_packets",
- "lro_bytes",
- "wqe_err",
- "mpwqe_filler",
- "mpwqe_frag",
- "buff_alloc_err",
-};
-
-struct mlx5e_rq_stats {
- u64 packets;
- u64 bytes;
- u64 csum_none;
- u64 csum_sw;
- u64 lro_packets;
- u64 lro_bytes;
- u64 wqe_err;
- u64 mpwqe_filler;
- u64 mpwqe_frag;
- u64 buff_alloc_err;
-#define NUM_RQ_STATS 10
-};
-
-static const char sq_stats_strings[][ETH_GSTRING_LEN] = {
- "packets",
- "bytes",
- "tso_packets",
- "tso_bytes",
- "tso_inner_packets",
- "tso_inner_bytes",
- "csum_offload_inner",
- "nop",
- "csum_offload_none",
- "stopped",
- "wake",
- "dropped",
-};
-
-struct mlx5e_sq_stats {
- /* commonly accessed in data path */
- u64 packets;
- u64 bytes;
- u64 tso_packets;
- u64 tso_bytes;
- u64 tso_inner_packets;
- u64 tso_inner_bytes;
- u64 csum_offload_inner;
- u64 nop;
- /* less likely accessed in data path */
- u64 csum_offload_none;
- u64 stopped;
- u64 wake;
- u64 dropped;
-#define NUM_SQ_STATS 12
-};
-
-struct mlx5e_stats {
- struct mlx5e_vport_stats vport;
- struct mlx5e_pport_stats pport;
- struct mlx5e_qcounter_stats qcnt;
-};
-
struct mlx5e_params {
u8 log_sq_size;
u8 rq_wq_type;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index 4077856..f1649d5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -166,6 +166,12 @@ static const struct {
};
#define MLX5E_NUM_Q_CNTRS(priv) (NUM_Q_COUNTERS * (!!priv->q_counter))
+#define MLX5E_NUM_RQ_STATS(priv) \
+ (NUM_RQ_STATS * priv->params.num_channels * \
+ test_bit(MLX5E_STATE_OPENED, &priv->state))
+#define MLX5E_NUM_SQ_STATS(priv) \
+ (NUM_SQ_STATS * priv->params.num_channels * priv->params.num_tc * \
+ test_bit(MLX5E_STATE_OPENED, &priv->state))
static int mlx5e_get_sset_count(struct net_device *dev, int sset)
{
@@ -173,21 +179,68 @@ static int mlx5e_get_sset_count(struct net_device *dev, int sset)
switch (sset) {
case ETH_SS_STATS:
- return NUM_VPORT_COUNTERS + NUM_PPORT_COUNTERS +
+ return NUM_SW_COUNTERS +
MLX5E_NUM_Q_CNTRS(priv) +
- priv->params.num_channels * NUM_RQ_STATS +
- priv->params.num_channels * priv->params.num_tc *
- NUM_SQ_STATS;
+ NUM_VPORT_COUNTERS + NUM_PPORT_COUNTERS +
+ MLX5E_NUM_RQ_STATS(priv) +
+ MLX5E_NUM_SQ_STATS(priv);
/* fallthrough */
default:
return -EOPNOTSUPP;
}
}
+static void mlx5e_fill_stats_strings(struct mlx5e_priv *priv, uint8_t *data)
+{
+ int i, j, tc, idx = 0;
+
+ /* SW counters */
+ for (i = 0; i < NUM_SW_COUNTERS; i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN, sw_stats_desc[i].name);
+
+ /* Q counters */
+ for (i = 0; i < MLX5E_NUM_Q_CNTRS(priv); i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN, q_stats_desc[i].name);
+
+ /* VPORT counters */
+ for (i = 0; i < NUM_VPORT_COUNTERS; i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN,
+ vport_stats_desc[i].name);
+
+ /* PPORT counters */
+ for (i = 0; i < NUM_PPORT_802_3_COUNTERS; i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN,
+ pport_802_3_stats_desc[i].name);
+
+ for (i = 0; i < NUM_PPORT_2863_COUNTERS; i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN,
+ pport_2863_stats_desc[i].name);
+
+ for (i = 0; i < NUM_PPORT_2819_COUNTERS; i++)
+ strcpy(data + (idx++) * ETH_GSTRING_LEN,
+ pport_2819_stats_desc[i].name);
+
+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
+ return;
+
+ /* per channel counters */
+ for (i = 0; i < priv->params.num_channels; i++)
+ for (j = 0; j < NUM_RQ_STATS; j++)
+ sprintf(data + (idx++) * ETH_GSTRING_LEN, "rx%d_%s", i,
+ rq_stats_desc[j].name);
+
+ for (tc = 0; tc < priv->params.num_tc; tc++)
+ for (i = 0; i < priv->params.num_channels; i++)
+ for (j = 0; j < NUM_SQ_STATS; j++)
+ sprintf(data + (idx++) * ETH_GSTRING_LEN,
+ "tx%d_%s",
+ priv->channeltc_to_txq_map[i][tc],
+ sq_stats_desc[j].name);
+}
+
static void mlx5e_get_strings(struct net_device *dev,
uint32_t stringset, uint8_t *data)
{
- int i, j, tc, idx = 0;
struct mlx5e_priv *priv = netdev_priv(dev);
switch (stringset) {
@@ -198,35 +251,7 @@ static void mlx5e_get_strings(struct net_device *dev,
break;
case ETH_SS_STATS:
- /* VPORT counters */
- for (i = 0; i < NUM_VPORT_COUNTERS; i++)
- strcpy(data + (idx++) * ETH_GSTRING_LEN,
- vport_strings[i]);
-
- /* Q counters */
- for (i = 0; i < MLX5E_NUM_Q_CNTRS(priv); i++)
- strcpy(data + (idx++) * ETH_GSTRING_LEN,
- qcounter_stats_strings[i]);
-
- /* PPORT counters */
- for (i = 0; i < NUM_PPORT_COUNTERS; i++)
- strcpy(data + (idx++) * ETH_GSTRING_LEN,
- pport_strings[i]);
-
- /* per channel counters */
- for (i = 0; i < priv->params.num_channels; i++)
- for (j = 0; j < NUM_RQ_STATS; j++)
- sprintf(data + (idx++) * ETH_GSTRING_LEN,
- "rx%d_%s", i, rq_stats_strings[j]);
-
- for (tc = 0; tc < priv->params.num_tc; tc++)
- for (i = 0; i < priv->params.num_channels; i++)
- for (j = 0; j < NUM_SQ_STATS; j++)
- sprintf(data +
- (idx++) * ETH_GSTRING_LEN,
- "tx%d_%s",
- priv->channeltc_to_txq_map[i][tc],
- sq_stats_strings[j]);
+ mlx5e_fill_stats_strings(priv, data);
break;
}
}
@@ -245,28 +270,45 @@ static void mlx5e_get_ethtool_stats(struct net_device *dev,
mlx5e_update_stats(priv);
mutex_unlock(&priv->state_lock);
- for (i = 0; i < NUM_VPORT_COUNTERS; i++)
- data[idx++] = ((u64 *)&priv->stats.vport)[i];
+ for (i = 0; i < NUM_SW_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_CPU(&priv->stats.sw,
+ sw_stats_desc, i);
for (i = 0; i < MLX5E_NUM_Q_CNTRS(priv); i++)
- data[idx++] = ((u32 *)&priv->stats.qcnt)[i];
+ data[idx++] = MLX5E_READ_CTR32_CPU(&priv->stats.qcnt,
+ q_stats_desc, i);
+
+ for (i = 0; i < NUM_VPORT_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_BE(priv->stats.vport.query_vport_out,
+ vport_stats_desc, i);
+
+ for (i = 0; i < NUM_PPORT_802_3_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.IEEE_802_3_counters,
+ pport_802_3_stats_desc, i);
- for (i = 0; i < NUM_PPORT_COUNTERS; i++)
- data[idx++] = be64_to_cpu(((__be64 *)&priv->stats.pport)[i]);
+ for (i = 0; i < NUM_PPORT_2863_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.RFC_2863_counters,
+ pport_2863_stats_desc, i);
+
+ for (i = 0; i < NUM_PPORT_2819_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.RFC_2819_counters,
+ pport_2819_stats_desc, i);
+
+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
+ return;
/* per channel counters */
for (i = 0; i < priv->params.num_channels; i++)
for (j = 0; j < NUM_RQ_STATS; j++)
- data[idx++] = !test_bit(MLX5E_STATE_OPENED,
- &priv->state) ? 0 :
- ((u64 *)&priv->channel[i]->rq.stats)[j];
+ data[idx++] =
+ MLX5E_READ_CTR64_CPU(&priv->channel[i]->rq.stats,
+ rq_stats_desc, j);
for (tc = 0; tc < priv->params.num_tc; tc++)
for (i = 0; i < priv->params.num_channels; i++)
for (j = 0; j < NUM_SQ_STATS; j++)
- data[idx++] = !test_bit(MLX5E_STATE_OPENED,
- &priv->state) ? 0 :
- ((u64 *)&priv->channel[i]->sq[tc].stats)[j];
+ data[idx++] = MLX5E_READ_CTR64_CPU(&priv->channel[i]->sq[tc].stats,
+ sq_stats_desc, j);
}
static void mlx5e_get_ringparam(struct net_device *dev,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 6270f8d..0c53236 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -91,96 +91,15 @@ static void mlx5e_update_carrier_work(struct work_struct *work)
mutex_unlock(&priv->state_lock);
}
-static void mlx5e_update_pport_counters(struct mlx5e_priv *priv)
-{
- struct mlx5_core_dev *mdev = priv->mdev;
- struct mlx5e_pport_stats *s = &priv->stats.pport;
- u32 *in;
- u32 *out;
- int sz = MLX5_ST_SZ_BYTES(ppcnt_reg);
-
- in = mlx5_vzalloc(sz);
- out = mlx5_vzalloc(sz);
- if (!in || !out)
- goto free_out;
-
- MLX5_SET(ppcnt_reg, in, local_port, 1);
-
- MLX5_SET(ppcnt_reg, in, grp, MLX5_IEEE_802_3_COUNTERS_GROUP);
- mlx5_core_access_reg(mdev, in, sz, out,
- sz, MLX5_REG_PPCNT, 0, 0);
- memcpy(s->IEEE_802_3_counters,
- MLX5_ADDR_OF(ppcnt_reg, out, counter_set),
- sizeof(s->IEEE_802_3_counters));
-
- MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2863_COUNTERS_GROUP);
- mlx5_core_access_reg(mdev, in, sz, out,
- sz, MLX5_REG_PPCNT, 0, 0);
- memcpy(s->RFC_2863_counters,
- MLX5_ADDR_OF(ppcnt_reg, out, counter_set),
- sizeof(s->RFC_2863_counters));
-
- MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP);
- mlx5_core_access_reg(mdev, in, sz, out,
- sz, MLX5_REG_PPCNT, 0, 0);
- memcpy(s->RFC_2819_counters,
- MLX5_ADDR_OF(ppcnt_reg, out, counter_set),
- sizeof(s->RFC_2819_counters));
-
-free_out:
- kvfree(in);
- kvfree(out);
-}
-
-static void mlx5e_update_q_counter(struct mlx5e_priv *priv)
-{
- struct mlx5e_qcounter_stats *qcnt = &priv->stats.qcnt;
-
- if (!priv->q_counter)
- return;
-
- mlx5_core_query_out_of_buffer(priv->mdev, priv->q_counter,
- &qcnt->rx_out_of_buffer);
-}
-
-void mlx5e_update_stats(struct mlx5e_priv *priv)
+static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
{
- struct mlx5_core_dev *mdev = priv->mdev;
- struct mlx5e_vport_stats *s = &priv->stats.vport;
+ struct mlx5e_sw_stats *s = &priv->stats.sw;
struct mlx5e_rq_stats *rq_stats;
struct mlx5e_sq_stats *sq_stats;
- u32 in[MLX5_ST_SZ_DW(query_vport_counter_in)];
- u32 *out;
- int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out);
- u64 tx_offload_none;
+ u64 tx_offload_none = 0;
int i, j;
- out = mlx5_vzalloc(outlen);
- if (!out)
- return;
-
- /* Collect firts the SW counters and then HW for consistency */
- s->rx_packets = 0;
- s->rx_bytes = 0;
- s->tx_packets = 0;
- s->tx_bytes = 0;
- s->tso_packets = 0;
- s->tso_bytes = 0;
- s->tso_inner_packets = 0;
- s->tso_inner_bytes = 0;
- s->tx_queue_stopped = 0;
- s->tx_queue_wake = 0;
- s->tx_queue_dropped = 0;
- s->tx_csum_inner = 0;
- tx_offload_none = 0;
- s->lro_packets = 0;
- s->lro_bytes = 0;
- s->rx_csum_none = 0;
- s->rx_csum_sw = 0;
- s->rx_wqe_err = 0;
- s->rx_mpwqe_filler = 0;
- s->rx_mpwqe_frag = 0;
- s->rx_buff_alloc_err = 0;
+ memset(s, 0, sizeof(*s));
for (i = 0; i < priv->params.num_channels; i++) {
rq_stats = &priv->channel[i]->rq.stats;
@@ -212,7 +131,19 @@ void mlx5e_update_stats(struct mlx5e_priv *priv)
}
}
- /* HW counters */
+ /* Update calculated offload counters */
+ s->tx_csum_offload = s->tx_packets - tx_offload_none - s->tx_csum_inner;
+ s->rx_csum_good = s->rx_packets - s->rx_csum_none -
+ s->rx_csum_sw;
+}
+
+static void mlx5e_update_vport_counters(struct mlx5e_priv *priv)
+{
+ int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out);
+ u32 *out = (u32 *)priv->stats.vport.query_vport_out;
+ u32 in[MLX5_ST_SZ_DW(query_vport_counter_in)];
+ struct mlx5_core_dev *mdev = priv->mdev;
+
memset(in, 0, sizeof(in));
MLX5_SET(query_vport_counter_in, in, opcode,
@@ -222,58 +153,56 @@ void mlx5e_update_stats(struct mlx5e_priv *priv)
memset(out, 0, outlen);
- if (mlx5_cmd_exec(mdev, in, sizeof(in), out, outlen))
+ mlx5_cmd_exec(mdev, in, sizeof(in), out, outlen);
+}
+
+static void mlx5e_update_pport_counters(struct mlx5e_priv *priv)
+{
+ struct mlx5e_pport_stats *pstats = &priv->stats.pport;
+ struct mlx5_core_dev *mdev = priv->mdev;
+ int sz = MLX5_ST_SZ_BYTES(ppcnt_reg);
+ void *out;
+ u32 *in;
+
+ in = mlx5_vzalloc(sz);
+ if (!in)
goto free_out;
-#define MLX5_GET_CTR(p, x) \
- MLX5_GET64(query_vport_counter_out, p, x)
-
- s->rx_error_packets =
- MLX5_GET_CTR(out, received_errors.packets);
- s->rx_error_bytes =
- MLX5_GET_CTR(out, received_errors.octets);
- s->tx_error_packets =
- MLX5_GET_CTR(out, transmit_errors.packets);
- s->tx_error_bytes =
- MLX5_GET_CTR(out, transmit_errors.octets);
-
- s->rx_unicast_packets =
- MLX5_GET_CTR(out, received_eth_unicast.packets);
- s->rx_unicast_bytes =
- MLX5_GET_CTR(out, received_eth_unicast.octets);
- s->tx_unicast_packets =
- MLX5_GET_CTR(out, transmitted_eth_unicast.packets);
- s->tx_unicast_bytes =
- MLX5_GET_CTR(out, transmitted_eth_unicast.octets);
-
- s->rx_multicast_packets =
- MLX5_GET_CTR(out, received_eth_multicast.packets);
- s->rx_multicast_bytes =
- MLX5_GET_CTR(out, received_eth_multicast.octets);
- s->tx_multicast_packets =
- MLX5_GET_CTR(out, transmitted_eth_multicast.packets);
- s->tx_multicast_bytes =
- MLX5_GET_CTR(out, transmitted_eth_multicast.octets);
-
- s->rx_broadcast_packets =
- MLX5_GET_CTR(out, received_eth_broadcast.packets);
- s->rx_broadcast_bytes =
- MLX5_GET_CTR(out, received_eth_broadcast.octets);
- s->tx_broadcast_packets =
- MLX5_GET_CTR(out, transmitted_eth_broadcast.packets);
- s->tx_broadcast_bytes =
- MLX5_GET_CTR(out, transmitted_eth_broadcast.octets);
+ MLX5_SET(ppcnt_reg, in, local_port, 1);
- /* Update calculated offload counters */
- s->tx_csum_offload = s->tx_packets - tx_offload_none - s->tx_csum_inner;
- s->rx_csum_good = s->rx_packets - s->rx_csum_none -
- s->rx_csum_sw;
+ out = pstats->IEEE_802_3_counters;
+ MLX5_SET(ppcnt_reg, in, grp, MLX5_IEEE_802_3_COUNTERS_GROUP);
+ mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
- mlx5e_update_pport_counters(priv);
- mlx5e_update_q_counter(priv);
+ out = pstats->RFC_2863_counters;
+ MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2863_COUNTERS_GROUP);
+ mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
+
+ out = pstats->RFC_2819_counters;
+ MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP);
+ mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
free_out:
- kvfree(out);
+ kvfree(in);
+}
+
+static void mlx5e_update_q_counter(struct mlx5e_priv *priv)
+{
+ struct mlx5e_qcounter_stats *qcnt = &priv->stats.qcnt;
+
+ if (!priv->q_counter)
+ return;
+
+ mlx5_core_query_out_of_buffer(priv->mdev, priv->q_counter,
+ &qcnt->rx_out_of_buffer);
+}
+
+void mlx5e_update_stats(struct mlx5e_priv *priv)
+{
+ mlx5e_update_sw_counters(priv);
+ mlx5e_update_q_counter(priv);
+ mlx5e_update_vport_counters(priv);
+ mlx5e_update_pport_counters(priv);
}
static void mlx5e_update_stats_work(struct work_struct *work)
@@ -2073,37 +2002,28 @@ static struct rtnl_link_stats64 *
mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats)
{
struct mlx5e_priv *priv = netdev_priv(dev);
+ struct mlx5e_sw_stats *sstats = &priv->stats.sw;
struct mlx5e_vport_stats *vstats = &priv->stats.vport;
struct mlx5e_pport_stats *pstats = &priv->stats.pport;
- stats->rx_packets = vstats->rx_packets;
- stats->rx_bytes = vstats->rx_bytes;
- stats->tx_packets = vstats->tx_packets;
- stats->tx_bytes = vstats->tx_bytes;
-
-#define PPCNT_GET_802_3_CTR(fld) \
- (MLX5_GET64(eth_802_3_cntrs_grp_data_layout, \
- pstats->IEEE_802_3_counters, fld##_high))
-
-#define PPCNT_GET_2863_CTR(fld) \
- (MLX5_GET64(eth_2863_cntrs_grp_data_layout, \
- pstats->RFC_2863_counters, fld##_high))
+ stats->rx_packets = sstats->rx_packets;
+ stats->rx_bytes = sstats->rx_bytes;
+ stats->tx_packets = sstats->tx_packets;
+ stats->tx_bytes = sstats->tx_bytes;
stats->rx_dropped = priv->stats.qcnt.rx_out_of_buffer;
- stats->tx_dropped = vstats->tx_queue_dropped;
+ stats->tx_dropped = sstats->tx_queue_dropped;
stats->rx_length_errors =
- PPCNT_GET_802_3_CTR(a_in_range_length_errors) +
- PPCNT_GET_802_3_CTR(a_out_of_range_length_field) +
- PPCNT_GET_802_3_CTR(a_frame_too_long_errors);
+ PPORT_802_3_GET(pstats, a_in_range_length_errors) +
+ PPORT_802_3_GET(pstats, a_out_of_range_length_field) +
+ PPORT_802_3_GET(pstats, a_frame_too_long_errors);
stats->rx_crc_errors =
- PPCNT_GET_802_3_CTR(a_frame_check_sequence_errors);
- stats->rx_frame_errors =
- PPCNT_GET_802_3_CTR(a_alignment_errors);
- stats->tx_aborted_errors =
- PPCNT_GET_2863_CTR(if_out_discards);
+ PPORT_802_3_GET(pstats, a_frame_check_sequence_errors);
+ stats->rx_frame_errors = PPORT_802_3_GET(pstats, a_alignment_errors);
+ stats->tx_aborted_errors = PPORT_2863_GET(pstats, if_out_discards);
stats->tx_carrier_errors =
- PPCNT_GET_802_3_CTR(a_symbol_error_during_carrier);
+ PPORT_802_3_GET(pstats, a_symbol_error_during_carrier);
stats->rx_errors = stats->rx_length_errors + stats->rx_crc_errors +
stats->rx_frame_errors;
stats->tx_errors = stats->tx_aborted_errors + stats->tx_carrier_errors;
@@ -2111,8 +2031,8 @@ mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats)
/* vport multicast also counts packets that are dropped due to steering
* or rx out of buffer
*/
- stats->multicast = vstats->rx_multicast_packets;
-
+ stats->multicast =
+ VPORT_COUNTER_GET(vstats, received_eth_multicast.packets);
return stats;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
new file mode 100644
index 0000000..116320d
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -0,0 +1,318 @@
+/*
+ * Copyright (c) 2015-2016, Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __MLX5_EN_STATS_H__
+#define __MLX5_EN_STATS_H__
+
+#define MLX5E_READ_CTR64_CPU(ptr, dsc, i) \
+ (*(u64 *)((char *)ptr + dsc[i].offset))
+#define MLX5E_READ_CTR64_BE(ptr, dsc, i) \
+ be64_to_cpu(*(__be64 *)((char *)ptr + dsc[i].offset))
+#define MLX5E_READ_CTR32_CPU(ptr, dsc, i) \
+ (*(u32 *)((char *)ptr + dsc[i].offset))
+#define MLX5E_READ_CTR32_BE(ptr, dsc, i) \
+ be64_to_cpu(*(__be32 *)((char *)ptr + dsc[i].offset))
+
+#define MLX5E_DECLARE_STAT(type, fld) #fld, offsetof(type, fld)
+
+struct counter_desc {
+ char name[ETH_GSTRING_LEN];
+ int offset; /* Byte offset */
+};
+
+struct mlx5e_sw_stats {
+ u64 rx_packets;
+ u64 rx_bytes;
+ u64 tx_packets;
+ u64 tx_bytes;
+ u64 tso_packets;
+ u64 tso_bytes;
+ u64 tso_inner_packets;
+ u64 tso_inner_bytes;
+ u64 lro_packets;
+ u64 lro_bytes;
+ u64 rx_csum_good;
+ u64 rx_csum_none;
+ u64 rx_csum_sw;
+ u64 tx_csum_offload;
+ u64 tx_csum_inner;
+ u64 tx_queue_stopped;
+ u64 tx_queue_wake;
+ u64 tx_queue_dropped;
+ u64 rx_wqe_err;
+ u64 rx_mpwqe_filler;
+ u64 rx_mpwqe_frag;
+ u64 rx_buff_alloc_err;
+};
+
+static const struct counter_desc sw_stats_desc[] = {
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tso_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tso_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tso_inner_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tso_inner_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, lro_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, lro_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_good) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_none) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_sw) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_offload) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_inner) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_queue_stopped) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_queue_wake) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_queue_dropped) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_wqe_err) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_mpwqe_filler) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_mpwqe_frag) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_buff_alloc_err) },
+};
+
+struct mlx5e_qcounter_stats {
+ u32 rx_out_of_buffer;
+};
+
+static const struct counter_desc q_stats_desc[] = {
+ { MLX5E_DECLARE_STAT(struct mlx5e_qcounter_stats, rx_out_of_buffer) },
+};
+
+#define VPORT_COUNTER_OFF(c) MLX5_BYTE_OFF(query_vport_counter_out, c)
+#define VPORT_COUNTER_GET(vstats, c) MLX5_GET64(query_vport_counter_out, \
+ vstats->query_vport_out, c)
+
+struct mlx5e_vport_stats {
+ __be64 query_vport_out[MLX5_ST_SZ_QW(query_vport_counter_out)];
+};
+
+static const struct counter_desc vport_stats_desc[] = {
+ { "rx_error_packets", VPORT_COUNTER_OFF(received_errors.packets) },
+ { "rx_error_bytes", VPORT_COUNTER_OFF(received_errors.octets) },
+ { "tx_error_packets", VPORT_COUNTER_OFF(transmit_errors.packets) },
+ { "tx_error_bytes", VPORT_COUNTER_OFF(transmit_errors.octets) },
+ { "rx_unicast_packets",
+ VPORT_COUNTER_OFF(received_eth_unicast.packets) },
+ { "rx_unicast_bytes", VPORT_COUNTER_OFF(received_eth_unicast.octets) },
+ { "tx_unicast_packets",
+ VPORT_COUNTER_OFF(transmitted_eth_unicast.packets) },
+ { "tx_unicast_bytes",
+ VPORT_COUNTER_OFF(transmitted_eth_unicast.octets) },
+ { "rx_multicast_packets",
+ VPORT_COUNTER_OFF(received_eth_multicast.packets) },
+ { "rx_multicast_bytes",
+ VPORT_COUNTER_OFF(received_eth_multicast.octets) },
+ { "tx_multicast_packets",
+ VPORT_COUNTER_OFF(transmitted_eth_multicast.packets) },
+ { "tx_multicast_bytes",
+ VPORT_COUNTER_OFF(transmitted_eth_multicast.octets) },
+ { "rx_broadcast_packets",
+ VPORT_COUNTER_OFF(received_eth_broadcast.packets) },
+ { "rx_broadcast_bytes",
+ VPORT_COUNTER_OFF(received_eth_broadcast.octets) },
+ { "tx_broadcast_packets",
+ VPORT_COUNTER_OFF(transmitted_eth_broadcast.packets) },
+ { "tx_broadcast_bytes",
+ VPORT_COUNTER_OFF(transmitted_eth_broadcast.octets) },
+};
+
+#define PPORT_802_3_OFF(c) \
+ MLX5_BYTE_OFF(ppcnt_reg, \
+ counter_set.eth_802_3_cntrs_grp_data_layout.c##_high)
+#define PPORT_802_3_GET(pstats, c) \
+ MLX5_GET64(ppcnt_reg, pstats->IEEE_802_3_counters, \
+ counter_set.eth_802_3_cntrs_grp_data_layout.c##_high)
+#define PPORT_2863_OFF(c) \
+ MLX5_BYTE_OFF(ppcnt_reg, \
+ counter_set.eth_2863_cntrs_grp_data_layout.c##_high)
+#define PPORT_2863_GET(pstats, c) \
+ MLX5_GET64(ppcnt_reg, pstats->RFC_2863_counters, \
+ counter_set.eth_2863_cntrs_grp_data_layout.c##_high)
+#define PPORT_2819_OFF(c) \
+ MLX5_BYTE_OFF(ppcnt_reg, \
+ counter_set.eth_2819_cntrs_grp_data_layout.c##_high)
+#define PPORT_2819_GET(pstats, c) \
+ MLX5_GET64(ppcnt_reg, pstats->RFC_2819_counters, \
+ counter_set.eth_2819_cntrs_grp_data_layout.c##_high)
+
+struct mlx5e_pport_stats {
+ __be64 IEEE_802_3_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
+ __be64 RFC_2863_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
+ __be64 RFC_2819_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
+};
+
+static const struct counter_desc pport_802_3_stats_desc[] = {
+ { "frames_tx", PPORT_802_3_OFF(a_frames_transmitted_ok) },
+ { "frames_rx", PPORT_802_3_OFF(a_frames_received_ok) },
+ { "check_seq_err", PPORT_802_3_OFF(a_frame_check_sequence_errors) },
+ { "alignment_err", PPORT_802_3_OFF(a_alignment_errors) },
+ { "octets_tx", PPORT_802_3_OFF(a_octets_transmitted_ok) },
+ { "octets_received", PPORT_802_3_OFF(a_octets_received_ok) },
+ { "multicast_xmitted", PPORT_802_3_OFF(a_multicast_frames_xmitted_ok) },
+ { "broadcast_xmitted", PPORT_802_3_OFF(a_broadcast_frames_xmitted_ok) },
+ { "multicast_rx", PPORT_802_3_OFF(a_multicast_frames_received_ok) },
+ { "broadcast_rx", PPORT_802_3_OFF(a_broadcast_frames_received_ok) },
+ { "in_range_len_errors", PPORT_802_3_OFF(a_in_range_length_errors) },
+ { "out_of_range_len", PPORT_802_3_OFF(a_out_of_range_length_field) },
+ { "too_long_errors", PPORT_802_3_OFF(a_frame_too_long_errors) },
+ { "symbol_err", PPORT_802_3_OFF(a_symbol_error_during_carrier) },
+ { "mac_control_tx", PPORT_802_3_OFF(a_mac_control_frames_transmitted) },
+ { "mac_control_rx", PPORT_802_3_OFF(a_mac_control_frames_received) },
+ { "unsupported_op_rx",
+ PPORT_802_3_OFF(a_unsupported_opcodes_received) },
+ { "pause_ctrl_rx", PPORT_802_3_OFF(a_pause_mac_ctrl_frames_received) },
+ { "pause_ctrl_tx",
+ PPORT_802_3_OFF(a_pause_mac_ctrl_frames_transmitted) },
+};
+
+static const struct counter_desc pport_2863_stats_desc[] = {
+ { "in_octets", PPORT_2863_OFF(if_in_octets) },
+ { "in_ucast_pkts", PPORT_2863_OFF(if_in_ucast_pkts) },
+ { "in_discards", PPORT_2863_OFF(if_in_discards) },
+ { "in_errors", PPORT_2863_OFF(if_in_errors) },
+ { "in_unknown_protos", PPORT_2863_OFF(if_in_unknown_protos) },
+ { "out_octets", PPORT_2863_OFF(if_out_octets) },
+ { "out_ucast_pkts", PPORT_2863_OFF(if_out_ucast_pkts) },
+ { "out_discards", PPORT_2863_OFF(if_out_discards) },
+ { "out_errors", PPORT_2863_OFF(if_out_errors) },
+ { "in_multicast_pkts", PPORT_2863_OFF(if_in_multicast_pkts) },
+ { "in_broadcast_pkts", PPORT_2863_OFF(if_in_broadcast_pkts) },
+ { "out_multicast_pkts", PPORT_2863_OFF(if_out_multicast_pkts) },
+ { "out_broadcast_pkts", PPORT_2863_OFF(if_out_broadcast_pkts) },
+};
+
+static const struct counter_desc pport_2819_stats_desc[] = {
+ { "drop_events", PPORT_2819_OFF(ether_stats_drop_events) },
+ { "octets", PPORT_2819_OFF(ether_stats_octets) },
+ { "pkts", PPORT_2819_OFF(ether_stats_pkts) },
+ { "broadcast_pkts", PPORT_2819_OFF(ether_stats_broadcast_pkts) },
+ { "multicast_pkts", PPORT_2819_OFF(ether_stats_multicast_pkts) },
+ { "crc_align_errors", PPORT_2819_OFF(ether_stats_crc_align_errors) },
+ { "undersize_pkts", PPORT_2819_OFF(ether_stats_undersize_pkts) },
+ { "oversize_pkts", PPORT_2819_OFF(ether_stats_oversize_pkts) },
+ { "fragments", PPORT_2819_OFF(ether_stats_fragments) },
+ { "jabbers", PPORT_2819_OFF(ether_stats_jabbers) },
+ { "collisions", PPORT_2819_OFF(ether_stats_collisions) },
+ { "p64octets", PPORT_2819_OFF(ether_stats_pkts64octets) },
+ { "p65to127octets", PPORT_2819_OFF(ether_stats_pkts65to127octets) },
+ { "p128to255octets", PPORT_2819_OFF(ether_stats_pkts128to255octets) },
+ { "p256to511octets", PPORT_2819_OFF(ether_stats_pkts256to511octets) },
+ { "p512to1023octets", PPORT_2819_OFF(ether_stats_pkts512to1023octets) },
+ { "p1024to1518octets",
+ PPORT_2819_OFF(ether_stats_pkts1024to1518octets) },
+ { "p1519to2047octets",
+ PPORT_2819_OFF(ether_stats_pkts1519to2047octets) },
+ { "p2048to4095octets",
+ PPORT_2819_OFF(ether_stats_pkts2048to4095octets) },
+ { "p4096to8191octets",
+ PPORT_2819_OFF(ether_stats_pkts4096to8191octets) },
+ { "p8192to10239octets",
+ PPORT_2819_OFF(ether_stats_pkts8192to10239octets) },
+};
+
+struct mlx5e_rq_stats {
+ u64 packets;
+ u64 bytes;
+ u64 csum_none;
+ u64 csum_sw;
+ u64 lro_packets;
+ u64 lro_bytes;
+ u64 wqe_err;
+ u64 mpwqe_filler;
+ u64 mpwqe_frag;
+ u64 buff_alloc_err;
+};
+
+static const struct counter_desc rq_stats_desc[] = {
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_none) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_sw) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, lro_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, lro_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, wqe_err) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, mpwqe_filler) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, mpwqe_frag) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, buff_alloc_err) },
+};
+
+struct mlx5e_sq_stats {
+ /* commonly accessed in data path */
+ u64 packets;
+ u64 bytes;
+ u64 tso_packets;
+ u64 tso_bytes;
+ u64 tso_inner_packets;
+ u64 tso_inner_bytes;
+ u64 csum_offload_inner;
+ u64 nop;
+ /* less likely accessed in data path */
+ u64 csum_offload_none;
+ u64 stopped;
+ u64 wake;
+ u64 dropped;
+};
+
+static const struct counter_desc sq_stats_desc[] = {
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, tso_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, tso_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, tso_inner_packets) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, tso_inner_bytes) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, csum_offload_inner) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, nop) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, csum_offload_none) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, stopped) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, wake) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sq_stats, dropped) },
+};
+
+#define NUM_SW_COUNTERS ARRAY_SIZE(sw_stats_desc)
+#define NUM_Q_COUNTERS ARRAY_SIZE(q_stats_desc)
+#define NUM_VPORT_COUNTERS ARRAY_SIZE(vport_stats_desc)
+#define NUM_PPORT_802_3_COUNTERS ARRAY_SIZE(pport_802_3_stats_desc)
+#define NUM_PPORT_2863_COUNTERS ARRAY_SIZE(pport_2863_stats_desc)
+#define NUM_PPORT_2819_COUNTERS ARRAY_SIZE(pport_2819_stats_desc)
+#define NUM_PPORT_COUNTERS (NUM_PPORT_802_3_COUNTERS + \
+ NUM_PPORT_2863_COUNTERS + \
+ NUM_PPORT_2819_COUNTERS)
+#define NUM_RQ_STATS ARRAY_SIZE(rq_stats_desc)
+#define NUM_SQ_STATS ARRAY_SIZE(sq_stats_desc)
+
+struct mlx5e_stats {
+ struct mlx5e_sw_stats sw;
+ struct mlx5e_qcounter_stats qcnt;
+ struct mlx5e_vport_stats vport;
+ struct mlx5e_pport_stats pport;
+};
+
+#endif /* __MLX5_EN_STATS_H__ */
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 03f8d71..8be44ca 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -59,6 +59,7 @@
#define MLX5_FLD_SZ_BYTES(typ, fld) (__mlx5_bit_sz(typ, fld) / 8)
#define MLX5_ST_SZ_BYTES(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 8)
#define MLX5_ST_SZ_DW(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 32)
+#define MLX5_ST_SZ_QW(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 64)
#define MLX5_UN_SZ_BYTES(typ) (sizeof(union mlx5_ifc_##typ##_bits) / 8)
#define MLX5_UN_SZ_DW(typ) (sizeof(union mlx5_ifc_##typ##_bits) / 32)
#define MLX5_BYTE_OFF(typ, fld) (__mlx5_bit_off(typ, fld) / 8)
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 11/11] net/mlx5e: Fix checksum handling for non-stripped vlan packets
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
Now as rx-vlan offload can be disabled, packets can be received
with vlan tag not stripped, which means is_first_ethertype_ip will
return false, for that we need to check if the hardware reported
csum OK so we will report CHECKSUM_UNNECESSARY for those packets.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 20 ++++++++++++++----
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 8 +++++-
include/linux/mlx5/device.h | 21 +++++++++++++++----
4 files changed, 38 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 6c9c10c..5bad17d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -109,6 +109,7 @@ static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
s->lro_bytes += rq_stats->lro_bytes;
s->rx_csum_none += rq_stats->csum_none;
s->rx_csum_sw += rq_stats->csum_sw;
+ s->rx_csum_inner += rq_stats->csum_inner;
s->rx_wqe_err += rq_stats->wqe_err;
s->rx_mpwqe_filler += rq_stats->mpwqe_filler;
s->rx_mpwqe_frag += rq_stats->mpwqe_frag;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 918b7c7..23adfe2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -543,16 +543,26 @@ static inline void mlx5e_handle_csum(struct net_device *netdev,
if (lro) {
skb->ip_summed = CHECKSUM_UNNECESSARY;
- } else if (likely(is_first_ethertype_ip(skb))) {
+ return;
+ }
+
+ if (is_first_ethertype_ip(skb)) {
skb->ip_summed = CHECKSUM_COMPLETE;
skb->csum = csum_unfold((__force __sum16)cqe->check_sum);
rq->stats.csum_sw++;
- } else {
- goto csum_none;
+ return;
}
- return;
-
+ if (likely((cqe->hds_ip_ext & CQE_L3_OK) &&
+ (cqe->hds_ip_ext & CQE_L4_OK))) {
+ skb->ip_summed = CHECKSUM_UNNECESSARY;
+ if (cqe_is_tunneled(cqe)) {
+ skb->csum_level = 1;
+ skb->encapsulation = 1;
+ rq->stats.csum_inner++;
+ }
+ return;
+ }
csum_none:
skb->ip_summed = CHECKSUM_NONE;
rq->stats.csum_none++;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index 7cd8cb4..115752b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -62,6 +62,7 @@ struct mlx5e_sw_stats {
u64 rx_csum_good;
u64 rx_csum_none;
u64 rx_csum_sw;
+ u64 rx_csum_inner;
u64 tx_csum_offload;
u64 tx_csum_inner;
u64 tx_queue_stopped;
@@ -90,6 +91,7 @@ static const struct counter_desc sw_stats_desc[] = {
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_good) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_none) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_sw) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_inner) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_offload) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_inner) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_queue_stopped) },
@@ -272,8 +274,9 @@ static const struct counter_desc pport_per_prio_pfc_stats_desc[] = {
struct mlx5e_rq_stats {
u64 packets;
u64 bytes;
- u64 csum_none;
u64 csum_sw;
+ u64 csum_inner;
+ u64 csum_none;
u64 lro_packets;
u64 lro_bytes;
u64 wqe_err;
@@ -285,8 +288,9 @@ struct mlx5e_rq_stats {
static const struct counter_desc rq_stats_desc[] = {
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, packets) },
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, bytes) },
- { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_none) },
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_sw) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_inner) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, csum_none) },
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, lro_packets) },
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, lro_bytes) },
{ MLX5E_DECLARE_STAT(struct mlx5e_rq_stats, wqe_err) },
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 942bcca..6bd429b 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -645,8 +645,9 @@ struct mlx5_err_cqe {
};
struct mlx5_cqe64 {
- u8 rsvd0[2];
- __be16 wqe_id;
+ u8 outer_l3_tunneled;
+ u8 rsvd0;
+ __be16 wqe_id;
u8 lro_tcppsh_abort_dupack;
u8 lro_min_ttl;
__be16 lro_tcp_win;
@@ -659,7 +660,7 @@ struct mlx5_cqe64 {
__be16 slid;
__be32 flags_rqpn;
u8 hds_ip_ext;
- u8 l4_hdr_type_etc;
+ u8 l4_l3_hdr_type;
__be16 vlan_info;
__be32 srqn; /* [31:24]: lro_num_seg, [23:0]: srqn */
__be32 imm_inval_pkey;
@@ -680,12 +681,22 @@ static inline int get_cqe_lro_tcppsh(struct mlx5_cqe64 *cqe)
static inline u8 get_cqe_l4_hdr_type(struct mlx5_cqe64 *cqe)
{
- return (cqe->l4_hdr_type_etc >> 4) & 0x7;
+ return (cqe->l4_l3_hdr_type >> 4) & 0x7;
+}
+
+static inline u8 get_cqe_l3_hdr_type(struct mlx5_cqe64 *cqe)
+{
+ return (cqe->l4_l3_hdr_type >> 2) & 0x3;
+}
+
+static inline u8 cqe_is_tunneled(struct mlx5_cqe64 *cqe)
+{
+ return cqe->outer_l3_tunneled & 0x1;
}
static inline int cqe_has_vlan(struct mlx5_cqe64 *cqe)
{
- return !!(cqe->l4_hdr_type_etc & 0x1);
+ return !!(cqe->l4_l3_hdr_type & 0x1);
}
static inline u64 get_cqe_ts(struct mlx5_cqe64 *cqe)
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 04/11] net/mlx5e: Add per priority group to PPort counters
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Expose counters providing information for each priority level (PCP) through
ethtool -S option and DCBNL.
This includes rx/tx bytes, frames, and pause counters.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c | 6 ++
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 51 ++++++++++++++++++-
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 9 ++++
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 31 ++++++++++++-
4 files changed, 93 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
index 3036f27..b2db180 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
@@ -174,8 +174,14 @@ static int mlx5e_dcbnl_ieee_getpfc(struct net_device *dev,
{
struct mlx5e_priv *priv = netdev_priv(dev);
struct mlx5_core_dev *mdev = priv->mdev;
+ struct mlx5e_pport_stats *pstats = &priv->stats.pport;
+ int i;
pfc->pfc_cap = mlx5_max_tc(mdev) + 1;
+ for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) {
+ pfc->requests[i] = PPORT_PER_PRIO_GET(pstats, i, tx_pause);
+ pfc->indications[i] = PPORT_PER_PRIO_GET(pstats, i, rx_pause);
+ }
return mlx5_query_port_pfc(mdev, &pfc->pfc_en, NULL);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index f1649d5..522d584 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -165,6 +165,18 @@ static const struct {
},
};
+static unsigned long mlx5e_query_pfc_combined(struct mlx5e_priv *priv)
+{
+ struct mlx5_core_dev *mdev = priv->mdev;
+ u8 pfc_en_tx;
+ u8 pfc_en_rx;
+ int err;
+
+ err = mlx5_query_port_pfc(mdev, &pfc_en_tx, &pfc_en_rx);
+
+ return err ? 0 : pfc_en_tx | pfc_en_rx;
+}
+
#define MLX5E_NUM_Q_CNTRS(priv) (NUM_Q_COUNTERS * (!!priv->q_counter))
#define MLX5E_NUM_RQ_STATS(priv) \
(NUM_RQ_STATS * priv->params.num_channels * \
@@ -172,6 +184,7 @@ static const struct {
#define MLX5E_NUM_SQ_STATS(priv) \
(NUM_SQ_STATS * priv->params.num_channels * priv->params.num_tc * \
test_bit(MLX5E_STATE_OPENED, &priv->state))
+#define MLX5E_NUM_PFC_COUNTERS(priv) hweight8(mlx5e_query_pfc_combined(priv))
static int mlx5e_get_sset_count(struct net_device *dev, int sset)
{
@@ -183,7 +196,8 @@ static int mlx5e_get_sset_count(struct net_device *dev, int sset)
MLX5E_NUM_Q_CNTRS(priv) +
NUM_VPORT_COUNTERS + NUM_PPORT_COUNTERS +
MLX5E_NUM_RQ_STATS(priv) +
- MLX5E_NUM_SQ_STATS(priv);
+ MLX5E_NUM_SQ_STATS(priv) +
+ MLX5E_NUM_PFC_COUNTERS(priv);
/* fallthrough */
default:
return -EOPNOTSUPP;
@@ -192,7 +206,8 @@ static int mlx5e_get_sset_count(struct net_device *dev, int sset)
static void mlx5e_fill_stats_strings(struct mlx5e_priv *priv, uint8_t *data)
{
- int i, j, tc, idx = 0;
+ int i, j, tc, prio, idx = 0;
+ unsigned long pfc_combined;
/* SW counters */
for (i = 0; i < NUM_SW_COUNTERS; i++)
@@ -220,6 +235,21 @@ static void mlx5e_fill_stats_strings(struct mlx5e_priv *priv, uint8_t *data)
strcpy(data + (idx++) * ETH_GSTRING_LEN,
pport_2819_stats_desc[i].name);
+ for (prio = 0; prio < NUM_PPORT_PRIO; prio++) {
+ for (i = 0; i < NUM_PPORT_PER_PRIO_TRAFFIC_COUNTERS; i++)
+ sprintf(data + (idx++) * ETH_GSTRING_LEN, "prio%d_%s",
+ prio,
+ pport_per_prio_traffic_stats_desc[i].name);
+ }
+
+ pfc_combined = mlx5e_query_pfc_combined(priv);
+ for_each_set_bit(prio, &pfc_combined, NUM_PPORT_PRIO) {
+ for (i = 0; i < NUM_PPORT_PER_PRIO_PFC_COUNTERS; i++) {
+ sprintf(data + (idx++) * ETH_GSTRING_LEN, "prio%d_%s",
+ prio, pport_per_prio_pfc_stats_desc[i].name);
+ }
+ }
+
if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
return;
@@ -260,7 +290,8 @@ static void mlx5e_get_ethtool_stats(struct net_device *dev,
struct ethtool_stats *stats, u64 *data)
{
struct mlx5e_priv *priv = netdev_priv(dev);
- int i, j, tc, idx = 0;
+ int i, j, tc, prio, idx = 0;
+ unsigned long pfc_combined;
if (!data)
return;
@@ -294,6 +325,20 @@ static void mlx5e_get_ethtool_stats(struct net_device *dev,
data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.RFC_2819_counters,
pport_2819_stats_desc, i);
+ for (prio = 0; prio < NUM_PPORT_PRIO; prio++) {
+ for (i = 0; i < NUM_PPORT_PER_PRIO_TRAFFIC_COUNTERS; i++)
+ data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.per_prio_counters[prio],
+ pport_per_prio_traffic_stats_desc, i);
+ }
+
+ pfc_combined = mlx5e_query_pfc_combined(priv);
+ for_each_set_bit(prio, &pfc_combined, NUM_PPORT_PRIO) {
+ for (i = 0; i < NUM_PPORT_PER_PRIO_PFC_COUNTERS; i++) {
+ data[idx++] = MLX5E_READ_CTR64_BE(&priv->stats.pport.per_prio_counters[prio],
+ pport_per_prio_pfc_stats_desc, i);
+ }
+ }
+
if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
return;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 0c53236..ef66ba6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -161,6 +161,7 @@ static void mlx5e_update_pport_counters(struct mlx5e_priv *priv)
struct mlx5e_pport_stats *pstats = &priv->stats.pport;
struct mlx5_core_dev *mdev = priv->mdev;
int sz = MLX5_ST_SZ_BYTES(ppcnt_reg);
+ int prio;
void *out;
u32 *in;
@@ -182,6 +183,14 @@ static void mlx5e_update_pport_counters(struct mlx5e_priv *priv)
MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP);
mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0);
+ MLX5_SET(ppcnt_reg, in, grp, MLX5_PER_PRIORITY_COUNTERS_GROUP);
+ for (prio = 0; prio < NUM_PPORT_PRIO; prio++) {
+ out = pstats->per_prio_counters[prio];
+ MLX5_SET(ppcnt_reg, in, prio_tc, prio);
+ mlx5_core_access_reg(mdev, in, sz, out, sz,
+ MLX5_REG_PPCNT, 0, 0);
+ }
+
free_out:
kvfree(in);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index 4f3a08d..de27eea 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -165,11 +165,19 @@ static const struct counter_desc vport_stats_desc[] = {
#define PPORT_2819_GET(pstats, c) \
MLX5_GET64(ppcnt_reg, pstats->RFC_2819_counters, \
counter_set.eth_2819_cntrs_grp_data_layout.c##_high)
+#define PPORT_PER_PRIO_OFF(c) \
+ MLX5_BYTE_OFF(ppcnt_reg, \
+ counter_set.eth_per_prio_grp_data_layout.c##_high)
+#define PPORT_PER_PRIO_GET(pstats, prio, c) \
+ MLX5_GET64(ppcnt_reg, pstats->per_prio_counters[prio], \
+ counter_set.eth_per_prio_grp_data_layout.c##_high)
+#define NUM_PPORT_PRIO 8
struct mlx5e_pport_stats {
__be64 IEEE_802_3_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
__be64 RFC_2863_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
__be64 RFC_2819_counters[MLX5_ST_SZ_QW(ppcnt_reg)];
+ __be64 per_prio_counters[NUM_PPORT_PRIO][MLX5_ST_SZ_QW(ppcnt_reg)];
};
static const struct counter_desc pport_802_3_stats_desc[] = {
@@ -241,6 +249,21 @@ static const struct counter_desc pport_2819_stats_desc[] = {
PPORT_2819_OFF(ether_stats_pkts8192to10239octets) },
};
+static const struct counter_desc pport_per_prio_traffic_stats_desc[] = {
+ { "rx_octets", PPORT_PER_PRIO_OFF(rx_octets) },
+ { "rx_frames", PPORT_PER_PRIO_OFF(rx_frames) },
+ { "tx_octets", PPORT_PER_PRIO_OFF(tx_octets) },
+ { "tx_frames", PPORT_PER_PRIO_OFF(tx_frames) },
+};
+
+static const struct counter_desc pport_per_prio_pfc_stats_desc[] = {
+ { "rx_pause", PPORT_PER_PRIO_OFF(rx_pause) },
+ { "rx_pause_duration", PPORT_PER_PRIO_OFF(rx_pause_duration) },
+ { "tx_pause", PPORT_PER_PRIO_OFF(tx_pause) },
+ { "tx_pause_duration", PPORT_PER_PRIO_OFF(tx_pause_duration) },
+ { "rx_pause_transition", PPORT_PER_PRIO_OFF(rx_pause_transition) },
+};
+
struct mlx5e_rq_stats {
u64 packets;
u64 bytes;
@@ -305,9 +328,15 @@ static const struct counter_desc sq_stats_desc[] = {
#define NUM_PPORT_802_3_COUNTERS ARRAY_SIZE(pport_802_3_stats_desc)
#define NUM_PPORT_2863_COUNTERS ARRAY_SIZE(pport_2863_stats_desc)
#define NUM_PPORT_2819_COUNTERS ARRAY_SIZE(pport_2819_stats_desc)
+#define NUM_PPORT_PER_PRIO_TRAFFIC_COUNTERS \
+ ARRAY_SIZE(pport_per_prio_traffic_stats_desc)
+#define NUM_PPORT_PER_PRIO_PFC_COUNTERS \
+ ARRAY_SIZE(pport_per_prio_pfc_stats_desc)
#define NUM_PPORT_COUNTERS (NUM_PPORT_802_3_COUNTERS + \
NUM_PPORT_2863_COUNTERS + \
- NUM_PPORT_2819_COUNTERS)
+ NUM_PPORT_2819_COUNTERS + \
+ NUM_PPORT_PER_PRIO_TRAFFIC_COUNTERS * \
+ NUM_PPORT_PRIO)
#define NUM_RQ_STATS ARRAY_SIZE(rq_stats_desc)
#define NUM_SQ_STATS ARRAY_SIZE(sq_stats_desc)
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 08/11] net/mlx5e: Add ethtool support for interface identify (LED blinking)
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Eugenia Emantayev, Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Add the needed hardware command and mlx5_ifc structs for managing LED
control.
Add set_phys_id ethtool callback to support ethtool -p flag.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 25 ++++++++++++++++++++
drivers/net/ethernet/mellanox/mlx5/core/port.c | 13 ++++++++++
include/linux/mlx5/driver.h | 1 +
include/linux/mlx5/port.h | 6 ++++
4 files changed, 45 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index 522d584..a2c444e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -1135,6 +1135,30 @@ static int mlx5e_set_wol(struct net_device *netdev, struct ethtool_wolinfo *wol)
return mlx5_set_port_wol(mdev, mlx5_wol_mode);
}
+static int mlx5e_set_phys_id(struct net_device *dev,
+ enum ethtool_phys_id_state state)
+{
+ struct mlx5e_priv *priv = netdev_priv(dev);
+ struct mlx5_core_dev *mdev = priv->mdev;
+ u16 beacon_duration;
+
+ if (!MLX5_CAP_GEN(mdev, beacon_led))
+ return -EOPNOTSUPP;
+
+ switch (state) {
+ case ETHTOOL_ID_ACTIVE:
+ beacon_duration = MLX5_BEACON_DURATION_INF;
+ break;
+ case ETHTOOL_ID_INACTIVE:
+ beacon_duration = MLX5_BEACON_DURATION_OFF;
+ break;
+ default:
+ return -EOPNOTSUPP;
+ }
+
+ return mlx5_set_port_beacon(mdev, beacon_duration);
+}
+
const struct ethtool_ops mlx5e_ethtool_ops = {
.get_drvinfo = mlx5e_get_drvinfo,
.get_link = ethtool_op_get_link,
@@ -1159,6 +1183,7 @@ const struct ethtool_ops mlx5e_ethtool_ops = {
.get_pauseparam = mlx5e_get_pauseparam,
.set_pauseparam = mlx5e_set_pauseparam,
.get_ts_info = mlx5e_get_ts_info,
+ .set_phys_id = mlx5e_set_phys_id,
.get_wol = mlx5e_get_wol,
.set_wol = mlx5e_set_wol,
};
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c
index c37740f..446549f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c
@@ -115,6 +115,19 @@ int mlx5_query_port_ptys(struct mlx5_core_dev *dev, u32 *ptys,
}
EXPORT_SYMBOL_GPL(mlx5_query_port_ptys);
+int mlx5_set_port_beacon(struct mlx5_core_dev *dev, u16 beacon_duration)
+{
+ u32 out[MLX5_ST_SZ_DW(mlcr_reg)];
+ u32 in[MLX5_ST_SZ_DW(mlcr_reg)];
+
+ memset(in, 0, sizeof(in));
+ MLX5_SET(mlcr_reg, in, local_port, 1);
+ MLX5_SET(mlcr_reg, in, beacon_duration, beacon_duration);
+
+ return mlx5_core_access_reg(dev, in, sizeof(in), out,
+ sizeof(out), MLX5_REG_MLCR, 0, 1);
+}
+
int mlx5_query_port_proto_cap(struct mlx5_core_dev *dev,
u32 *proto_cap, int proto_mask)
{
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 497a4db..2e8758d 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -116,6 +116,7 @@ enum {
MLX5_REG_PMLP = 0, /* TBD */
MLX5_REG_NODE_DESC = 0x6001,
MLX5_REG_HOST_ENDIANNESS = 0x7004,
+ MLX5_REG_MLCR = 0x902b,
};
enum {
diff --git a/include/linux/mlx5/port.h b/include/linux/mlx5/port.h
index 577e953..a364ab1 100644
--- a/include/linux/mlx5/port.h
+++ b/include/linux/mlx5/port.h
@@ -35,6 +35,11 @@
#include <linux/mlx5/driver.h>
+enum mlx5_beacon_duration {
+ MLX5_BEACON_DURATION_OFF = 0x0,
+ MLX5_BEACON_DURATION_INF = 0xffff,
+};
+
int mlx5_set_port_caps(struct mlx5_core_dev *dev, u8 port_num, u32 caps);
int mlx5_query_port_ptys(struct mlx5_core_dev *dev, u32 *ptys,
int ptys_size, int proto_mask, u8 local_port);
@@ -53,6 +58,7 @@ int mlx5_set_port_admin_status(struct mlx5_core_dev *dev,
enum mlx5_port_status status);
int mlx5_query_port_admin_status(struct mlx5_core_dev *dev,
enum mlx5_port_status *status);
+int mlx5_set_port_beacon(struct mlx5_core_dev *dev, u16 beacon_duration);
int mlx5_set_port_mtu(struct mlx5_core_dev *dev, int mtu, u8 port);
void mlx5_query_port_max_mtu(struct mlx5_core_dev *dev, int *max_mtu, u8 port);
--
1.7.1
^ permalink raw reply related
* [PATCH net-next V1 10/11] net/mlx5e: Add ethtool support for rxvlan-offload (vlan stripping)
From: Saeed Mahameed @ 2016-04-24 19:51 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Or Gerlitz, Tal Alon, Eran Ben Elisha, Gal Pressman,
Saeed Mahameed
In-Reply-To: <1461527516-29290-1-git-send-email-saeedm@mellanox.com>
From: Gal Pressman <galp@mellanox.com>
Use ethtool -K <interface> rxvlan <on/off> to enable/disable
C-TAG vlan stripping by hardware.
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 3 +
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 74 ++++++++++++++++++++-
include/linux/mlx5/driver.h | 4 +
3 files changed, 78 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index e903eff..8abc289 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -166,6 +166,7 @@ struct mlx5e_params {
u8 rss_hfunc;
u8 toeplitz_hash_key[40];
u32 indirection_rqt[MLX5E_INDIR_RQT_SIZE];
+ bool vlan_strip_disable;
#ifdef CONFIG_MLX5_CORE_EN_DCB
struct ieee_ets ets;
#endif
@@ -575,6 +576,8 @@ int mlx5e_vlan_rx_kill_vid(struct net_device *dev, __always_unused __be16 proto,
void mlx5e_enable_vlan_filter(struct mlx5e_priv *priv);
void mlx5e_disable_vlan_filter(struct mlx5e_priv *priv);
+int mlx5e_modify_rqs_vsd(struct mlx5e_priv *priv, bool vsd);
+
int mlx5e_redirect_rqt(struct mlx5e_priv *priv, enum mlx5e_rqt_ix rqt_ix);
void mlx5e_build_tir_ctx_hash(void *tirc, struct mlx5e_priv *priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index ad0cb4a..6c9c10c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -388,6 +388,7 @@ static int mlx5e_enable_rq(struct mlx5e_rq *rq, struct mlx5e_rq_param *param)
MLX5_SET(rqc, rqc, cqn, rq->cq.mcq.cqn);
MLX5_SET(rqc, rqc, state, MLX5_RQC_STATE_RST);
MLX5_SET(rqc, rqc, flush_in_error_en, 1);
+ MLX5_SET(rqc, rqc, vsd, priv->params.vlan_strip_disable);
MLX5_SET(wq, wq, log_wq_pg_sz, rq->wq_ctrl.buf.page_shift -
MLX5_ADAPTER_PAGE_SHIFT);
MLX5_SET64(wq, wq, dbr_addr, rq->wq_ctrl.db.dma);
@@ -402,7 +403,8 @@ static int mlx5e_enable_rq(struct mlx5e_rq *rq, struct mlx5e_rq_param *param)
return err;
}
-static int mlx5e_modify_rq(struct mlx5e_rq *rq, int curr_state, int next_state)
+static int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state,
+ int next_state)
{
struct mlx5e_channel *c = rq->channel;
struct mlx5e_priv *priv = c->priv;
@@ -430,6 +432,36 @@ static int mlx5e_modify_rq(struct mlx5e_rq *rq, int curr_state, int next_state)
return err;
}
+static int mlx5e_modify_rq_vsd(struct mlx5e_rq *rq, bool vsd)
+{
+ struct mlx5e_channel *c = rq->channel;
+ struct mlx5e_priv *priv = c->priv;
+ struct mlx5_core_dev *mdev = priv->mdev;
+
+ void *in;
+ void *rqc;
+ int inlen;
+ int err;
+
+ inlen = MLX5_ST_SZ_BYTES(modify_rq_in);
+ in = mlx5_vzalloc(inlen);
+ if (!in)
+ return -ENOMEM;
+
+ rqc = MLX5_ADDR_OF(modify_rq_in, in, ctx);
+
+ MLX5_SET(modify_rq_in, in, rq_state, MLX5_RQC_STATE_RDY);
+ MLX5_SET64(modify_rq_in, in, modify_bitmask, MLX5_RQ_BITMASK_VSD);
+ MLX5_SET(rqc, rqc, vsd, vsd);
+ MLX5_SET(rqc, rqc, state, MLX5_RQC_STATE_RDY);
+
+ err = mlx5_core_modify_rq(mdev, rq->rqn, in, inlen);
+
+ kvfree(in);
+
+ return err;
+}
+
static void mlx5e_disable_rq(struct mlx5e_rq *rq)
{
mlx5_core_destroy_rq(rq->priv->mdev, rq->rqn);
@@ -468,7 +500,7 @@ static int mlx5e_open_rq(struct mlx5e_channel *c,
if (err)
goto err_destroy_rq;
- err = mlx5e_modify_rq(rq, MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY);
+ err = mlx5e_modify_rq_state(rq, MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY);
if (err)
goto err_disable_rq;
@@ -493,7 +525,7 @@ static void mlx5e_close_rq(struct mlx5e_rq *rq)
clear_bit(MLX5E_RQ_STATE_POST_WQES_ENABLE, &rq->state);
napi_synchronize(&rq->channel->napi); /* prevent mlx5e_post_rx_wqes */
- mlx5e_modify_rq(rq, MLX5_RQC_STATE_RDY, MLX5_RQC_STATE_ERR);
+ mlx5e_modify_rq_state(rq, MLX5_RQC_STATE_RDY, MLX5_RQC_STATE_ERR);
while (!mlx5_wq_ll_is_empty(&rq->wq))
msleep(20);
@@ -1963,6 +1995,23 @@ static void mlx5e_destroy_tirs(struct mlx5e_priv *priv)
mlx5e_destroy_tir(priv, i);
}
+int mlx5e_modify_rqs_vsd(struct mlx5e_priv *priv, bool vsd)
+{
+ int err = 0;
+ int i;
+
+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
+ return 0;
+
+ for (i = 0; i < priv->params.num_channels; i++) {
+ err = mlx5e_modify_rq_vsd(&priv->channel[i]->rq, vsd);
+ if (err)
+ return err;
+ }
+
+ return 0;
+}
+
static int mlx5e_setup_tc(struct net_device *netdev, u8 tc)
{
struct mlx5e_priv *priv = netdev_priv(netdev);
@@ -2147,6 +2196,23 @@ static int set_feature_rx_all(struct net_device *netdev, bool enable)
return mlx5_set_port_fcs(mdev, !enable);
}
+static int set_feature_rx_vlan(struct net_device *netdev, bool enable)
+{
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+ int err;
+
+ mutex_lock(&priv->state_lock);
+
+ priv->params.vlan_strip_disable = !enable;
+ err = mlx5e_modify_rqs_vsd(priv, !enable);
+ if (err)
+ priv->params.vlan_strip_disable = enable;
+
+ mutex_unlock(&priv->state_lock);
+
+ return err;
+}
+
static int mlx5e_handle_feature(struct net_device *netdev,
netdev_features_t wanted_features,
netdev_features_t feature,
@@ -2184,6 +2250,8 @@ static int mlx5e_set_features(struct net_device *netdev,
set_feature_tc_num_filters);
err |= mlx5e_handle_feature(netdev, features, NETIF_F_RXALL,
set_feature_rx_all);
+ err |= mlx5e_handle_feature(netdev, features, NETIF_F_HW_VLAN_CTAG_RX,
+ set_feature_rx_vlan);
return err ? -EINVAL : 0;
}
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 1a17067..2cc5e9f 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -46,6 +46,10 @@
#include <linux/mlx5/doorbell.h>
enum {
+ MLX5_RQ_BITMASK_VSD = 1 << 1,
+};
+
+enum {
MLX5_BOARD_ID_LEN = 64,
MLX5_MAX_NAME_LEN = 16,
};
--
1.7.1
^ permalink raw reply related
* Re: linux-next: zillions of lockdep whinges in include/net/sock.h:1408
From: Eric Dumazet @ 2016-04-24 19:55 UTC (permalink / raw)
To: David Miller; +Cc: hannes, Valdis.Kletnieks, netdev, linux-kernel
In-Reply-To: <20160424.145421.128023840849290353.davem@davemloft.net>
On Sun, 2016-04-24 at 14:54 -0400, David Miller wrote:
> From: Hannes Frederic Sowa <hannes@stressinduktion.org>
> Date: Sun, 24 Apr 2016 20:48:24 +0200
>
> > Eric's patch is worth to apply anyway, but I am not sure if it solves
> > the (fundamental) problem. I couldn't reproduce it with the exact next-
> > tag provided in the initial mail. All other reports also only happend
> > with linux-next and not net-next.
>
> Ok, Eric please submit it formally.
>
> Thanks!
Yes, I will first test it, and will take Hannes suggestion as well.
^ permalink raw reply
* Re: linux-next: zillions of lockdep whinges in include/net/sock.h:1408
From: Valdis.Kletnieks @ 2016-04-24 19:56 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Hannes Frederic Sowa, David Miller, netdev, linux-kernel
In-Reply-To: <1461527202.5535.1.camel@edumazet-glaptop3.roam.corp.google.com>
[-- Attachment #1: Type: text/plain, Size: 853 bytes --]
On Sun, 24 Apr 2016 12:46:42 -0700, Eric Dumazet said:
> >>> + return !debug_locks ||
> >>> + lockdep_is_held(&sk->sk_lock) ||
> Issue here is that once lockdep detected a problem (not necessarily in
> net/ tree btw), your helper always 'detect' a problem, since lockdep
> automatically disables itself.
"D'Oh!" -- H. Simpson
I thought this patch looked suspect, but couldn't put my finger on it. The
reason why I got like 41,000 of them is because I built a kernel that has
lockdep enabled, but I have an out-of-tree module that doesn't init something,
so I get this:
[ 48.898156] INFO: trying to register non-static key.
[ 48.898157] the code is fine but needs lockdep annotation.
[ 48.898157] turning off the locking correctness validator.
After which point, even with this patch, every time through it's still going to
explode.
[-- Attachment #2: Type: application/pgp-signature, Size: 848 bytes --]
^ permalink raw reply
* [PATCH net] RDMA/nes: don't leak skb if carrier down
From: Florian Westphal @ 2016-04-24 20:18 UTC (permalink / raw)
To: netdev-u79uwXL29TY76Z2rM5mHXA
Cc: Florian Westphal, linux-rdma-u79uwXL29TY76Z2rM5mHXA
Alternatively one could free the skb, OTOH I don't think this test is
useful so just remove it.
Cc: <linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Signed-off-by: Florian Westphal <fw-HFFVJYpyMKqzQB+pC5nmwQ@public.gmane.org>
---
Noticed this while working on the TX_LOCKED removal.
diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 3ea9e05..9291453 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -500,9 +500,6 @@ static int nes_netdev_start_xmit(struct sk_buff *skb, struct net_device *netdev)
* skb_shinfo(skb)->nr_frags, skb_is_gso(skb));
*/
- if (!netif_carrier_ok(netdev))
- return NETDEV_TX_OK;
-
if (netif_queue_stopped(netdev))
return NETDEV_TX_BUSY;
--
2.7.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH 0/2] sh_eth: couple of software reset bit cleanups
From: Sergei Shtylyov @ 2016-04-24 20:42 UTC (permalink / raw)
To: netdev; +Cc: linux-renesas-soc
Hello.
Here's a set of 2 patches against DaveM's 'net-next.git' repo. We clean up
the use of the software reset bits...
[1/2] sh_eth: use EDMR_SRST_GETHER in sh_eth_check_reset()
[2/2] sh_eth: rename ARSTR register bit
MBR, Sergei
^ permalink raw reply
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