* Re: [PATCH net-next 6/7] bnxt_en: Implement xmit_more.
From: Mintz, Yuval @ 2017-05-30 3:28 UTC (permalink / raw)
To: Michael Chan, davem@davemloft.net; +Cc: netdev@vger.kernel.org
In-Reply-To: <1496099170-701-7-git-send-email-michael.chan@broadcom.com>
> - bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
> + if (!skb->xmit_more)
> + bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
DQL can prevent you from getting the xmit_more marked SKB,
so you should probably also check for netif_xmit_stopped().
^ permalink raw reply
* Re: [PATCH net-next 0/7] bnxt_en: Misc. updates for net-next.
From: David Miller @ 2017-05-30 3:24 UTC (permalink / raw)
To: michael.chan; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
From: Michael Chan <michael.chan@broadcom.com>
Date: Mon, 29 May 2017 19:06:03 -0400
> The 1st 2 patches add short firmware message support for new VF devices.
> The 3rd patch adds a pci shutdown callback for the RDMA driver for proper
> shutdown. The next 3 patches improve the doorbell operations by
> elimiating the double doorbell workaround on newer chips, and by adding
> xmit_more support. The last patch adds a parameter to bnxt_set_dflt_rings().
Series applied, but where is the change that sets the new ulp_shutdown
method pointer to something?
Thanks.
^ permalink raw reply
* Re: [PATCH 0/4] Netfilter fixes for net
From: David Miller @ 2017-05-30 3:20 UTC (permalink / raw)
To: pablo; +Cc: netfilter-devel, netdev
In-Reply-To: <1496057672-6751-1-git-send-email-pablo@netfilter.org>
From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Mon, 29 May 2017 13:34:28 +0200
> Hi David,
>
> The following patchset contains Netfilter fixes for your net tree,
> they are:
>
> 1) Conntrack SCTP CRC32c checksum mangling may operate on non-linear
> skbuff, patch from Davide Caratti.
>
> 2) nf_tables rb-tree set backend does not handle element re-addition
> after deletion in the same transaction, leading to infinite loop.
>
> 3) Atomically unclear the IPS_SRC_NAT_DONE_BIT on nat module removal,
> from Liping Zhang.
>
> 4) Conntrack hashtable resizing while ctnetlink dump is progress leads
> to a dead reference to released objects in the lists, also from
> Liping.
>
> You can pull these changes from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
Pulled, thanks a lot Pablo.
^ permalink raw reply
* Re: [PATCH net-next 1/3] netvsc: optimize calculation of number of slots
From: David Miller @ 2017-05-30 0:42 UTC (permalink / raw)
To: stephen; +Cc: netdev, sthemmin
In-Reply-To: <20170529183159.22205-1-sthemmin@microsoft.com>
Stephen, please provide proper header "[PATCH net-next 0/3] " postings
with your patch series, so we know at a high level what this patch
series is doing, how it is doing it, and why it is doing it that
way.
I've been quite liberal with your patch postings in the past, hoping
you would catch on, but now I'm going to simply enforce what I ask
of every other patch series submitter.
Thanks.
^ permalink raw reply
* [PATCH net-next 6/7] bnxt_en: Implement xmit_more.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
Do not write the TX doorbell if skb->xmit_more is set unless the TX
queue is full.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index d7c1295..2c6af31 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -463,13 +463,17 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev)
prod = NEXT_TX(prod);
txr->tx_prod = prod;
- bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
+ if (!skb->xmit_more)
+ bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
tx_done:
mmiowb();
if (unlikely(bnxt_tx_avail(bp, txr) <= MAX_SKB_FRAGS + 1)) {
+ if (skb->xmit_more && !tx_buf->is_push)
+ bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
+
netif_tx_stop_queue(txq);
/* netif_tx_stop_queue() must be done before checking
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 7/7] bnxt_en: Pass in sh parameter to bnxt_set_dflt_rings().
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
In the existing code, the local variable sh is hardcoded to true to
calculate default rings for shared ring configuration. It is better
to have the caller determine the value of sh.
Reported-by: Gustavo A. R. Silva <garsilva@embeddedor.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 2c6af31..954758f 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -7567,10 +7567,9 @@ static int bnxt_get_dflt_rings(struct bnxt *bp, int *max_rx, int *max_tx,
return rc;
}
-static int bnxt_set_dflt_rings(struct bnxt *bp)
+static int bnxt_set_dflt_rings(struct bnxt *bp, bool sh)
{
int dflt_rings, max_rx_rings, max_tx_rings, rc;
- bool sh = true;
if (sh)
bp->flags |= BNXT_FLAG_SHARED_RINGS;
@@ -7749,7 +7748,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
bnxt_set_tpa_flags(bp);
bnxt_set_ring_params(bp);
bnxt_set_max_func_irqs(bp, max_irqs);
- rc = bnxt_set_dflt_rings(bp);
+ rc = bnxt_set_dflt_rings(bp, true);
if (rc) {
netdev_err(bp->dev, "Not enough rings available.\n");
rc = -ENOMEM;
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 5/7] bnxt_en: Optimize doorbell write operations for newer chips.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
Older chips require the doorbells to be written twice, but newer chips
do not. Add a new common function bnxt_db_write() to write all
doorbells appropriately depending on the chip. Eliminating the extra
doorbell on newer chips has a significant performance improvement
on pktgen.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 30 +++++++++--------------
drivers/net/ethernet/broadcom/bnxt/bnxt.h | 9 +++++++
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 3 +--
drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c | 2 +-
4 files changed, 23 insertions(+), 21 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 427db49..d7c1295 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -463,8 +463,7 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev)
prod = NEXT_TX(prod);
txr->tx_prod = prod;
- writel(DB_KEY_TX | prod, txr->tx_doorbell);
- writel(DB_KEY_TX | prod, txr->tx_doorbell);
+ bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | prod);
tx_done:
@@ -1779,8 +1778,7 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget)
/* Sync BD data before updating doorbell */
wmb();
- writel(DB_KEY_TX | prod, db);
- writel(DB_KEY_TX | prod, db);
+ bnxt_db_write(bp, db, DB_KEY_TX | prod);
}
cpr->cp_raw_cons = raw_cons;
@@ -1796,14 +1794,10 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget)
if (event & BNXT_RX_EVENT) {
struct bnxt_rx_ring_info *rxr = bnapi->rx_ring;
- writel(DB_KEY_RX | rxr->rx_prod, rxr->rx_doorbell);
- writel(DB_KEY_RX | rxr->rx_prod, rxr->rx_doorbell);
- if (event & BNXT_AGG_EVENT) {
- writel(DB_KEY_RX | rxr->rx_agg_prod,
- rxr->rx_agg_doorbell);
- writel(DB_KEY_RX | rxr->rx_agg_prod,
- rxr->rx_agg_doorbell);
- }
+ bnxt_db_write(bp, rxr->rx_doorbell, DB_KEY_RX | rxr->rx_prod);
+ if (event & BNXT_AGG_EVENT)
+ bnxt_db_write(bp, rxr->rx_agg_doorbell,
+ DB_KEY_RX | rxr->rx_agg_prod);
}
return rx_pkts;
}
@@ -1863,13 +1857,11 @@ static int bnxt_poll_nitroa0(struct napi_struct *napi, int budget)
cpr->cp_raw_cons = raw_cons;
BNXT_CP_DB(cpr->cp_doorbell, cpr->cp_raw_cons);
- writel(DB_KEY_RX | rxr->rx_prod, rxr->rx_doorbell);
- writel(DB_KEY_RX | rxr->rx_prod, rxr->rx_doorbell);
+ bnxt_db_write(bp, rxr->rx_doorbell, DB_KEY_RX | rxr->rx_prod);
- if (event & BNXT_AGG_EVENT) {
- writel(DB_KEY_RX | rxr->rx_agg_prod, rxr->rx_agg_doorbell);
- writel(DB_KEY_RX | rxr->rx_agg_prod, rxr->rx_agg_doorbell);
- }
+ if (event & BNXT_AGG_EVENT)
+ bnxt_db_write(bp, rxr->rx_agg_doorbell,
+ DB_KEY_RX | rxr->rx_agg_prod);
if (!bnxt_has_work(bp, cpr) && rx_pkts < budget) {
napi_complete_done(napi, rx_pkts);
@@ -7714,6 +7706,8 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
bp->gro_func = bnxt_gro_func_5730x;
if (BNXT_CHIP_P4_PLUS(bp))
bp->gro_func = bnxt_gro_func_5731x;
+ else
+ bp->flags |= BNXT_FLAG_DOUBLE_DB;
rc = bnxt_hwrm_func_drv_rgtr(bp);
if (rc)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index c59b2cd..5984423 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -1022,6 +1022,7 @@ struct bnxt {
#define BNXT_FLAG_FW_LLDP_AGENT 0x80000
#define BNXT_FLAG_MULTI_HOST 0x100000
#define BNXT_FLAG_SHORT_CMD 0x200000
+ #define BNXT_FLAG_DOUBLE_DB 0x400000
#define BNXT_FLAG_CHIP_NITRO_A0 0x1000000
#define BNXT_FLAG_ALL_CONFIG_FEATS (BNXT_FLAG_TPA | \
@@ -1254,6 +1255,14 @@ static inline u32 bnxt_tx_avail(struct bnxt *bp, struct bnxt_tx_ring_info *txr)
((txr->tx_prod - txr->tx_cons) & bp->tx_ring_mask);
}
+/* For TX and RX ring doorbells */
+static inline void bnxt_db_write(struct bnxt *bp, void __iomem *db, u32 val)
+{
+ writel(val, db);
+ if (bp->flags & BNXT_FLAG_DOUBLE_DB)
+ writel(val, db);
+}
+
extern const u16 bnxt_lhint_arr[];
int bnxt_alloc_rx_data(struct bnxt *bp, struct bnxt_rx_ring_info *rxr,
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 11ddf0a..fd11815 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -2376,8 +2376,7 @@ static int bnxt_run_loopback(struct bnxt *bp)
/* Sync BD data before updating doorbell */
wmb();
- writel(DB_KEY_TX | txr->tx_prod, txr->tx_doorbell);
- writel(DB_KEY_TX | txr->tx_prod, txr->tx_doorbell);
+ bnxt_db_write(bp, txr->tx_doorbell, DB_KEY_TX | txr->tx_prod);
rc = bnxt_poll_loopback(bp, pkt_size);
dma_unmap_single(&bp->pdev->dev, map, pkt_size, PCI_DMA_TODEVICE);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
index 9dae327..8ce793a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
@@ -63,7 +63,7 @@ void bnxt_tx_int_xdp(struct bnxt *bp, struct bnxt_napi *bnapi, int nr_pkts)
tx_buf = &txr->tx_buf_ring[last_tx_cons];
rx_prod = tx_buf->rx_prod;
}
- writel(DB_KEY_RX | rx_prod, rxr->rx_doorbell);
+ bnxt_db_write(bp, rxr->rx_doorbell, DB_KEY_RX | rx_prod);
}
/* returns the following:
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 2/7] bnxt_en: Add PCI IDs for BCM57454 VF devices.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
From: Deepak Khungar <deepak.khungar@broadcom.com>
Signed-off-by: Deepak Khungar <deepak.khungar@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 47bc4f3..dc10a9a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -175,6 +175,8 @@ enum board_idx {
{ PCI_VDEVICE(BROADCOM, 0x16f1), .driver_data = BCM57452 },
{ PCI_VDEVICE(BROADCOM, 0x1614), .driver_data = BCM57454 },
#ifdef CONFIG_BNXT_SRIOV
+ { PCI_VDEVICE(BROADCOM, 0x1606), .driver_data = NETXTREME_E_VF },
+ { PCI_VDEVICE(BROADCOM, 0x1609), .driver_data = NETXTREME_E_VF },
{ PCI_VDEVICE(BROADCOM, 0x16c1), .driver_data = NETXTREME_E_VF },
{ PCI_VDEVICE(BROADCOM, 0x16cb), .driver_data = NETXTREME_C_VF },
{ PCI_VDEVICE(BROADCOM, 0x16d3), .driver_data = NETXTREME_E_VF },
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 4/7] bnxt_en: Add additional chip ID definitions.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
Add additional chip definitions and macros for all supported chips.
Add a new macro BNXT_CHIP_P4_PLUS for the newer generation of chips and
use the macro to properly determine the features supported by these
newer chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 6 ++----
drivers/net/ethernet/broadcom/bnxt/bnxt.h | 29 +++++++++++++++++++++++++----
2 files changed, 27 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 63d23b6..427db49 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -7712,7 +7712,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
init_waitqueue_head(&bp->sriov_cfg_wait);
#endif
bp->gro_func = bnxt_gro_func_5730x;
- if (BNXT_CHIP_NUM_57X1X(bp->chip_num))
+ if (BNXT_CHIP_P4_PLUS(bp))
bp->gro_func = bnxt_gro_func_5731x;
rc = bnxt_hwrm_func_drv_rgtr(bp);
@@ -7763,9 +7763,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
VNIC_RSS_CFG_REQ_HASH_TYPE_TCP_IPV4 |
VNIC_RSS_CFG_REQ_HASH_TYPE_IPV6 |
VNIC_RSS_CFG_REQ_HASH_TYPE_TCP_IPV6;
- if (!BNXT_CHIP_NUM_57X0X(bp->chip_num) &&
- !BNXT_CHIP_TYPE_NITRO_A0(bp) &&
- bp->hwrm_spec_code >= 0x10501) {
+ if (BNXT_CHIP_P4_PLUS(bp) && bp->hwrm_spec_code >= 0x10501) {
bp->flags |= BNXT_FLAG_UDP_RSS_CAP;
bp->rss_hash_cfg |= VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV4 |
VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV6;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index cb56678..c59b2cd 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -938,31 +938,45 @@ struct bnxt {
#define CHIP_NUM_57402 0x16d0
#define CHIP_NUM_57404 0x16d1
#define CHIP_NUM_57406 0x16d2
+#define CHIP_NUM_57407 0x16d5
#define CHIP_NUM_57311 0x16ce
#define CHIP_NUM_57312 0x16cf
#define CHIP_NUM_57314 0x16df
+#define CHIP_NUM_57317 0x16e0
#define CHIP_NUM_57412 0x16d6
#define CHIP_NUM_57414 0x16d7
#define CHIP_NUM_57416 0x16d8
#define CHIP_NUM_57417 0x16d9
+#define CHIP_NUM_57412L 0x16da
+#define CHIP_NUM_57414L 0x16db
+
+#define CHIP_NUM_5745X 0xd730
#define BNXT_CHIP_NUM_5730X(chip_num) \
((chip_num) >= CHIP_NUM_57301 && \
(chip_num) <= CHIP_NUM_57304)
#define BNXT_CHIP_NUM_5740X(chip_num) \
- ((chip_num) >= CHIP_NUM_57402 && \
- (chip_num) <= CHIP_NUM_57406)
+ (((chip_num) >= CHIP_NUM_57402 && \
+ (chip_num) <= CHIP_NUM_57406) || \
+ (chip_num) == CHIP_NUM_57407)
#define BNXT_CHIP_NUM_5731X(chip_num) \
((chip_num) == CHIP_NUM_57311 || \
(chip_num) == CHIP_NUM_57312 || \
- (chip_num) == CHIP_NUM_57314)
+ (chip_num) == CHIP_NUM_57314 || \
+ (chip_num) == CHIP_NUM_57317)
#define BNXT_CHIP_NUM_5741X(chip_num) \
((chip_num) >= CHIP_NUM_57412 && \
- (chip_num) <= CHIP_NUM_57417)
+ (chip_num) <= CHIP_NUM_57414L)
+
+#define BNXT_CHIP_NUM_58700(chip_num) \
+ ((chip_num) == CHIP_NUM_58700)
+
+#define BNXT_CHIP_NUM_5745X(chip_num) \
+ ((chip_num) == CHIP_NUM_5745X)
#define BNXT_CHIP_NUM_57X0X(chip_num) \
(BNXT_CHIP_NUM_5730X(chip_num) || BNXT_CHIP_NUM_5740X(chip_num))
@@ -1022,6 +1036,13 @@ struct bnxt {
#define BNXT_CHIP_TYPE_NITRO_A0(bp) ((bp)->flags & BNXT_FLAG_CHIP_NITRO_A0)
#define BNXT_RX_PAGE_MODE(bp) ((bp)->flags & BNXT_FLAG_RX_PAGE_MODE)
+/* Chip class phase 4 and later */
+#define BNXT_CHIP_P4_PLUS(bp) \
+ (BNXT_CHIP_NUM_57X1X((bp)->chip_num) || \
+ BNXT_CHIP_NUM_5745X((bp)->chip_num) || \
+ (BNXT_CHIP_NUM_58700((bp)->chip_num) && \
+ !BNXT_CHIP_TYPE_NITRO_A0(bp)))
+
struct bnxt_en_dev *edev;
struct bnxt_en_dev * (*ulp_probe)(struct net_device *);
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 1/7] bnxt_en: Support for Short Firmware Message
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
From: Deepak Khungar <deepak.khungar@broadcom.com>
The new short message format is used on the new BCM57454 VFs. Each
firmware message is a fixed 16-byte message sent using the standard
firmware communication channel. The short message has a DMA address
pointing to the legacy long firmware message.
Signed-off-by: Deepak Khungar <deepak.khungar@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 66 ++++++++++++++++++++++++++++++-
drivers/net/ethernet/broadcom/bnxt/bnxt.h | 4 ++
2 files changed, 69 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 69b6829..47bc4f3 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -2868,6 +2868,32 @@ static int bnxt_alloc_hwrm_resources(struct bnxt *bp)
return 0;
}
+static void bnxt_free_hwrm_short_cmd_req(struct bnxt *bp)
+{
+ if (bp->hwrm_short_cmd_req_addr) {
+ struct pci_dev *pdev = bp->pdev;
+
+ dma_free_coherent(&pdev->dev, BNXT_HWRM_MAX_REQ_LEN,
+ bp->hwrm_short_cmd_req_addr,
+ bp->hwrm_short_cmd_req_dma_addr);
+ bp->hwrm_short_cmd_req_addr = NULL;
+ }
+}
+
+static int bnxt_alloc_hwrm_short_cmd_req(struct bnxt *bp)
+{
+ struct pci_dev *pdev = bp->pdev;
+
+ bp->hwrm_short_cmd_req_addr =
+ dma_alloc_coherent(&pdev->dev, BNXT_HWRM_MAX_REQ_LEN,
+ &bp->hwrm_short_cmd_req_dma_addr,
+ GFP_KERNEL);
+ if (!bp->hwrm_short_cmd_req_addr)
+ return -ENOMEM;
+
+ return 0;
+}
+
static void bnxt_free_stats(struct bnxt *bp)
{
u32 size, i;
@@ -3215,16 +3241,41 @@ static int bnxt_hwrm_do_send_msg(struct bnxt *bp, void *msg, u32 msg_len,
__le32 *resp_len, *valid;
u16 cp_ring_id, len = 0;
struct hwrm_err_output *resp = bp->hwrm_cmd_resp_addr;
+ u16 max_req_len = BNXT_HWRM_MAX_REQ_LEN;
req->seq_id = cpu_to_le16(bp->hwrm_cmd_seq++);
memset(resp, 0, PAGE_SIZE);
cp_ring_id = le16_to_cpu(req->cmpl_ring);
intr_process = (cp_ring_id == INVALID_HW_RING_ID) ? 0 : 1;
+ if (bp->flags & BNXT_FLAG_SHORT_CMD) {
+ void *short_cmd_req = bp->hwrm_short_cmd_req_addr;
+ struct hwrm_short_input short_input = {0};
+
+ memcpy(short_cmd_req, req, msg_len);
+ memset(short_cmd_req + msg_len, 0, BNXT_HWRM_MAX_REQ_LEN -
+ msg_len);
+
+ short_input.req_type = req->req_type;
+ short_input.signature =
+ cpu_to_le16(SHORT_REQ_SIGNATURE_SHORT_CMD);
+ short_input.size = cpu_to_le16(msg_len);
+ short_input.req_addr =
+ cpu_to_le64(bp->hwrm_short_cmd_req_dma_addr);
+
+ data = (u32 *)&short_input;
+ msg_len = sizeof(short_input);
+
+ /* Sync memory write before updating doorbell */
+ wmb();
+
+ max_req_len = BNXT_HWRM_SHORT_REQ_LEN;
+ }
+
/* Write request msg to hwrm channel */
__iowrite32_copy(bp->bar0, data, msg_len / 4);
- for (i = msg_len; i < BNXT_HWRM_MAX_REQ_LEN; i += 4)
+ for (i = msg_len; i < max_req_len; i += 4)
writel(0, bp->bar0 + i);
/* currently supports only one outstanding message */
@@ -4662,6 +4713,7 @@ static int bnxt_hwrm_ver_get(struct bnxt *bp)
int rc;
struct hwrm_ver_get_input req = {0};
struct hwrm_ver_get_output *resp = bp->hwrm_cmd_resp_addr;
+ u32 dev_caps_cfg;
bp->hwrm_max_req_len = HWRM_MAX_REQ_LEN;
bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_VER_GET, -1, -1);
@@ -4699,6 +4751,11 @@ static int bnxt_hwrm_ver_get(struct bnxt *bp)
!resp->chip_metal)
bp->flags |= BNXT_FLAG_CHIP_NITRO_A0;
+ dev_caps_cfg = le32_to_cpu(resp->dev_caps_cfg);
+ if ((dev_caps_cfg & VER_GET_RESP_DEV_CAPS_CFG_SHORT_CMD_SUPPORTED) &&
+ (dev_caps_cfg & VER_GET_RESP_DEV_CAPS_CFG_SHORT_CMD_REQUIRED))
+ bp->flags |= BNXT_FLAG_SHORT_CMD;
+
hwrm_ver_get_exit:
mutex_unlock(&bp->hwrm_cmd_lock);
return rc;
@@ -7357,6 +7414,7 @@ static void bnxt_remove_one(struct pci_dev *pdev)
bnxt_clear_int_mode(bp);
bnxt_hwrm_func_drv_unrgtr(bp);
bnxt_free_hwrm_resources(bp);
+ bnxt_free_hwrm_short_cmd_req(bp);
bnxt_ethtool_free(bp);
bnxt_dcb_free(bp);
kfree(bp->edev);
@@ -7607,6 +7665,12 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
if (rc)
goto init_err_pci_clean;
+ if (bp->flags & BNXT_FLAG_SHORT_CMD) {
+ rc = bnxt_alloc_hwrm_short_cmd_req(bp);
+ if (rc)
+ goto init_err_pci_clean;
+ }
+
rc = bnxt_hwrm_func_reset(bp);
if (rc)
goto init_err_pci_clean;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 3ef42db..cb56678 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -500,6 +500,7 @@ struct rx_tpa_end_cmp_ext {
#define NEXT_CMP(idx) RING_CMP(ADV_RAW_CMP(idx, 1))
#define BNXT_HWRM_MAX_REQ_LEN (bp->hwrm_max_req_len)
+#define BNXT_HWRM_SHORT_REQ_LEN sizeof(struct hwrm_short_input)
#define DFLT_HWRM_CMD_TIMEOUT 500
#define HWRM_CMD_TIMEOUT (bp->hwrm_cmd_timeout)
#define HWRM_RESET_TIMEOUT ((HWRM_CMD_TIMEOUT) * 4)
@@ -1006,6 +1007,7 @@ struct bnxt {
#define BNXT_FLAG_RX_PAGE_MODE 0x40000
#define BNXT_FLAG_FW_LLDP_AGENT 0x80000
#define BNXT_FLAG_MULTI_HOST 0x100000
+ #define BNXT_FLAG_SHORT_CMD 0x200000
#define BNXT_FLAG_CHIP_NITRO_A0 0x1000000
#define BNXT_FLAG_ALL_CONFIG_FEATS (BNXT_FLAG_TPA | \
@@ -1106,6 +1108,8 @@ struct bnxt {
u32 hwrm_spec_code;
u16 hwrm_cmd_seq;
u32 hwrm_intr_seq_id;
+ void *hwrm_short_cmd_req_addr;
+ dma_addr_t hwrm_short_cmd_req_dma_addr;
void *hwrm_cmd_resp_addr;
dma_addr_t hwrm_cmd_resp_dma_addr;
void *hwrm_dbg_resp_addr;
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 3/7] bnxt_en: Add a callback to inform RDMA driver during PCI shutdown.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1496099170-701-1-git-send-email-michael.chan@broadcom.com>
When bnxt_en gets a PCI shutdown call, we need to have a new callback
to inform the RDMA driver to do proper shutdown and removal.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 1 +
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 19 +++++++++++++++++++
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 2 ++
3 files changed, 22 insertions(+)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index dc10a9a..63d23b6 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -7837,6 +7837,7 @@ static void bnxt_shutdown(struct pci_dev *pdev)
dev_close(dev);
if (system_state == SYSTEM_POWER_OFF) {
+ bnxt_ulp_shutdown(bp);
bnxt_clear_int_mode(bp);
pci_wake_from_d3(pdev, bp->wol);
pci_set_power_state(pdev, PCI_D3hot);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c
index 8b7464b..77da75a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c
@@ -266,6 +266,25 @@ void bnxt_ulp_sriov_cfg(struct bnxt *bp, int num_vfs)
}
}
+void bnxt_ulp_shutdown(struct bnxt *bp)
+{
+ struct bnxt_en_dev *edev = bp->edev;
+ struct bnxt_ulp_ops *ops;
+ int i;
+
+ if (!edev)
+ return;
+
+ for (i = 0; i < BNXT_MAX_ULP; i++) {
+ struct bnxt_ulp *ulp = &edev->ulp_tbl[i];
+
+ ops = rtnl_dereference(ulp->ulp_ops);
+ if (!ops || !ops->ulp_shutdown)
+ continue;
+ ops->ulp_shutdown(ulp->handle);
+ }
+}
+
void bnxt_ulp_async_events(struct bnxt *bp, struct hwrm_async_event_cmpl *cmpl)
{
u16 event_id = le16_to_cpu(cmpl->event_id);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h
index 74f816e..d247106 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h
@@ -26,6 +26,7 @@ struct bnxt_ulp_ops {
void (*ulp_stop)(void *);
void (*ulp_start)(void *);
void (*ulp_sriov_config)(void *, int);
+ void (*ulp_shutdown)(void *);
};
struct bnxt_msix_entry {
@@ -87,6 +88,7 @@ static inline bool bnxt_ulp_registered(struct bnxt_en_dev *edev, int ulp_id)
void bnxt_ulp_stop(struct bnxt *bp);
void bnxt_ulp_start(struct bnxt *bp);
void bnxt_ulp_sriov_cfg(struct bnxt *bp, int num_vfs);
+void bnxt_ulp_shutdown(struct bnxt *bp);
void bnxt_ulp_async_events(struct bnxt *bp, struct hwrm_async_event_cmpl *cmpl);
struct bnxt_en_dev *bnxt_ulp_probe(struct net_device *dev);
--
1.8.3.1
^ permalink raw reply related
* [PATCH net-next 0/7] bnxt_en: Misc. updates for net-next.
From: Michael Chan @ 2017-05-29 23:06 UTC (permalink / raw)
To: davem; +Cc: netdev
The 1st 2 patches add short firmware message support for new VF devices.
The 3rd patch adds a pci shutdown callback for the RDMA driver for proper
shutdown. The next 3 patches improve the doorbell operations by
elimiating the double doorbell workaround on newer chips, and by adding
xmit_more support. The last patch adds a parameter to bnxt_set_dflt_rings().
Deepak Khungar (2):
bnxt_en: Support for Short Firmware Message
bnxt_en: Add PCI IDs for BCM57454 VF devices.
Michael Chan (5):
bnxt_en: Add a callback to inform RDMA driver during PCI shutdown.
bnxt_en: Add additional chip ID definitions.
bnxt_en: Optimize doorbell write operations for newer chips.
bnxt_en: Implement xmit_more.
bnxt_en: Pass in sh parameter to bnxt_set_dflt_rings().
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 114 +++++++++++++++++-----
drivers/net/ethernet/broadcom/bnxt/bnxt.h | 42 +++++++-
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 3 +-
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 19 ++++
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 2 +
drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c | 2 +-
6 files changed, 149 insertions(+), 33 deletions(-)
--
1.8.3.1
^ permalink raw reply
* _Зgравствуйтe! Bас интeрeсуют клиeнтскиe 6азы gанныx?
From: l.leontjevscky@yandex.ru @ 2017-05-29 22:55 UTC (permalink / raw)
_3дpaвcтвyйте! Bac uнтеpеcyют kлuентckuе 6aзы дaнных?
^ permalink raw reply
* Re: [patch net-next 0/9] mlxsw: Support firmware flash
From: Jakub Kicinski @ 2017-05-29 21:40 UTC (permalink / raw)
To: Or Gerlitz
Cc: Yotam Gigi, David Miller, Yuval.Mintz, Jiri Pirko,
Linux Netdev List, Ido Schimmel, mlxsw, ben
In-Reply-To: <CAJ3xEMgBAWrM2jKRmS0yYi94wEo-8FaV=G24_Z2=emkySV=f4A@mail.gmail.com>
[swapping Ben's email]
On Mon, 29 May 2017 11:52:27 +0300, Or Gerlitz wrote:
> On Mon, May 29, 2017 at 3:17 AM, Jakub Kicinski <kubakici@wp.pl> wrote:
> > On Sun, 28 May 2017 10:26:49 +0300, Yotam Gigi wrote:
>
> >> This problem is even more relevant in the Mellanox HCA driver team, which would
> >> like to use that code in order to burn the HCA firmware, but not intend to
> >> trigger it on boot time, which means that must have a way for the user to
> >> trigger it.
>
> > What would the requirements for the HCA team be? Is it about loading
> > different code or loading HW settings?
>
> For the NIC, as of the (happily growing) large open-systems by nature
> install base, we
> can't effort for the driver not to load when the currently burned fw
> isn't the latest or
> the system doesn't have the latest libfirmware clone. We do have to
> keep the hassle of
> compatibility with older FWs, etc. As part of the driver/FW API
> design we use cap bits
> for that matter, and never ask/branch on specific FW or HW versions.
>
> We do want to let user work with the kernel w.o the need to have them
> install MLNX or anyone's
> tool suit and hence the large value we find in the ethtool flashing of
> FW image (the patches
> are ready...)
Thanks for the explanations Or and Yotam! That does sound like a plain
version upgrade problem.
^ permalink raw reply
* Re: [PATCH] net: dsa: mv88e6xxx: Add missing static to stub functions
From: Arnd Bergmann @ 2017-05-29 21:06 UTC (permalink / raw)
To: Andrew Lunn
Cc: Vivien Didelot, Florian Fainelli, David S. Miller, Networking,
Linux Kernel Mailing List
In-Reply-To: <20170529153030.GB21315@lunn.ch>
On Mon, May 29, 2017 at 5:30 PM, Andrew Lunn <andrew@lunn.ch> wrote:
> On Mon, May 29, 2017 at 02:56:01PM +0200, Arnd Bergmann wrote:
>> 'static' was not enough, the helpers must be 'static inline'
>>
>> net/dsa/mv88e6xxx/global2.h:123:12: error: 'mv88e6xxx_g2_misc_4_bit_port' defined but not used [-Werror=unused-function]
>> net/dsa/mv88e6xxx/global2.h:117:12: error: 'mv88e6xxx_g2_pvt_write' defined but not used [-Werror=unused-function]
>>
>> Fixes: c21fbe29f858 ("net: dsa: mv88e6xxx: Add missing static to stub functions")
>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>
> Hi Arnd
>
> Dave merged my fix yesterday.
Hi Andrew,
The only commit I see from you in the net/master branch is the partial fix
I referenced in the 'fixes' line (c21fbe29f858), from all I can tell, my fix is
still needed on top. Can you check again?
Arnd
^ permalink raw reply
* Re: [PATCH v4 2/2] dt-bindings: net: Add Cortina device tree bindings
From: Florian Fainelli @ 2017-05-29 20:30 UTC (permalink / raw)
To: Bogdan Purcareata, andrew, netdev, devicetree, linux-kernel
In-Reply-To: <1496049091-4637-3-git-send-email-bogdan.purcareata@nxp.com>
Le 05/29/17 à 02:11, Bogdan Purcareata a écrit :
> Add device tree description info for Cortina 10G phy devices.
>
> Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* Re: [PATCH v4 1/2] net: phy: Add Cortina CS4340 driver
From: Florian Fainelli @ 2017-05-29 20:28 UTC (permalink / raw)
To: Bogdan Purcareata, andrew-g2DYL2Zd6BY,
netdev-u79uwXL29TY76Z2rM5mHXA, devicetree-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1496049091-4637-2-git-send-email-bogdan.purcareata-3arQi8VN3Tc@public.gmane.org>
Le 05/29/17 à 02:11, Bogdan Purcareata a écrit :
> Add basic support for Cortina PHY drivers. Support only CS4340 for now.
> The phys are not compatible with IEEE 802.3 clause 22/45 registers.
>
> Implement proper read_status support. The generic 10G phy driver causes
> bus register access errors.
>
> The driver should be described using the "ethernet-phy-id" device tree
> compatible.
>
> Signed-off-by: Bogdan Purcareata <bogdan.purcareata-3arQi8VN3Tc@public.gmane.org>
Reviewed-by: Florian Fainelli <f.fainelli-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
--
Florian
--
To unsubscribe from this list: send the line "unsubscribe devicetree" 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
* _Зqpавствуйтe! Bас uнтepeсуют клueнтскue 6азы qанных?
From: senior.balmont@yandex.ru @ 2017-05-29 19:44 UTC (permalink / raw)
_3дpавсmвуйmе! Вас uнmеpесуюm клuенmскuе базы данных?
^ permalink raw reply
* Re: vxlan: use after free error
From: Roopa Prabhu @ 2017-05-29 18:37 UTC (permalink / raw)
To: Mark Bloch
Cc: davem@davemloft.net, Jiri Benc, pravin shelar, Alexander Duyck,
Nicolas Dichtel, netdev@vger.kernel.org, Balki Raman
In-Reply-To: <b04f5f7e-dca1-b219-eb52-03f58fd3823e@mellanox.com>
On Sun, May 28, 2017 at 11:28 PM, Mark Bloch <markb@mellanox.com> wrote:
> Hi Roopa,
>
> On 29/05/2017 05:50, Roopa Prabhu wrote:
>> On Sun, May 28, 2017 at 3:49 AM, Mark Bloch <markb@mellanox.com> wrote:
[snip]
>>
>> From: Balakrishnan Raman <ramanb@cumulusnetworks.com>
>>
>> Date: Sun, 28 May 2017 19:34:25 -0700
>>
>> Subject: [PATCH net-next] vxlan: remove vxlan device from socket's device
>>
>> list in vxlan_stop
>>
>>
>> Signed-off-by: Balakrishnan Raman <ramanb@cumulusnetworks.com>
>>
>> Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
>>
>> ---
>>
>> drivers/net/vxlan.c | 13 ++++++++++++-
>>
>> 1 file changed, 12 insertions(+), 1 deletion(-)
>>
>>
>> diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
>>
>> index 328b471..26dac52 100644
>>
>> --- a/drivers/net/vxlan.c
>>
>> +++ b/drivers/net/vxlan.c
>>
>> @@ -2352,6 +2352,16 @@ static void vxlan_vs_add_dev(struct vxlan_sock
>> *vs, struct vxlan_dev *vxlan)
>>
>> spin_unlock(&vn->sock_lock);
>>
>> }
>>
>>
>> +static void vxlan_vs_del_dev(struct vxlan_dev *vxlan)
>>
>> +{
>>
>> + struct vxlan_net *vn = net_generic(vxlan->net, vxlan_net_id);
>>
>> +
>>
>> + spin_lock(&vn->sock_lock);
>>
>> + if (!hlist_unhashed(&vxlan->hlist))
>
> We are calling this from vxlan_stop(), which means we did vxlan_open(), there
> we do vxlan_sock_add() and added the vxlan to the socket. so the check
> is unnecessary and we can just call hlist_del_init_rcu() directly.
What you say looks correct..., but does not hurt to leave this check in there..
given rest of the changes you are proposing below.
Looking at git blame, this check was added for OVS in dellink...but it
could have been because
it was being called before stop in dellink.
>
>>
>> + hlist_del_init_rcu(&vxlan->hlist);
>>
>> + spin_unlock(&vn->sock_lock);
>>
>> +}
>>
>> +
>>
>> /* Setup stats when device is created */
>>
>> static int vxlan_init(struct net_device *dev)
>>
>> {
>>
>> @@ -2443,6 +2453,7 @@ static int vxlan_stop(struct net_device *dev)
>>
>> del_timer_sync(&vxlan->age_timer);
>>
>>
>> vxlan_flush(vxlan, false);
>>
>> + vxlan_vs_del_dev(vxlan);
>
> In my patch I've added the code inside vxlan_sock_release()
> after we do:
> rcu_assign_pointer(vxlan->vn6_sock, NULL);
> rcu_assign_pointer(vxlan->vn4_sock, NULL);
> synchronize_net();
> this way we make sure we are done handling packets that may access the
> vxlan struct. seems cleaner to me, what do you think?
sure, that works. But, lets keep the function vxlan_vs_del_dev(vxlan)
for symmetry with
vxlan_vs_add_dev(vxlan) and call it from vxlan_sock_release().
>
>>
>> vxlan_sock_release(vxlan);
>>
>>
>> return ret;
>>
>>
>> @@ -3292,7 +3303,7 @@ static void vxlan_dellink(struct net_device
>> *dev, struct list_head *head)
>>
>>
>> spin_lock(&vn->sock_lock);
>>
>> if (!hlist_unhashed(&vxlan->hlist))
>>
>> - hlist_del_rcu(&vxlan->hlist);
>>
>> + hlist_del_init_rcu(&vxlan->hlist);
>>
>> spin_unlock(&vn->sock_lock);
>>
>
> in vxlan_dellink() we call unregister_netdevice_queue(). which means if
> the interface is open, the kernel would stop it. (and we'll hit the code in vxlan_stop()).
> We can remove this entire block of code it seems.
That seems right. It does look redundant if we hit the same code via
vxlan_stop during dellink.
This code is also hit via the OVS path, and i don't see a problem with
your changes and analysis but i am not too familiar with the ovs call
path. I see that the relevant developers are CC'ed.
thanks.
^ permalink raw reply
* [PATCH net-next 3/3] netvsc: use typed pointer for internal state
From: Stephen Hemminger @ 2017-05-29 18:31 UTC (permalink / raw)
To: davem; +Cc: netdev, Stephen Hemminger
In-Reply-To: <20170529183159.22205-1-sthemmin@microsoft.com>
The element netvsc_device:extension is always a pointer to RNDIS
information.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
---
drivers/net/hyperv/hyperv_net.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/net/hyperv/hyperv_net.h b/drivers/net/hyperv/hyperv_net.h
index 262b2ea576a3..f82d54e0208c 100644
--- a/drivers/net/hyperv/hyperv_net.h
+++ b/drivers/net/hyperv/hyperv_net.h
@@ -763,8 +763,7 @@ struct netvsc_device {
refcount_t sc_offered;
- /* Holds rndis device info */
- void *extension;
+ struct rndis_device *extension;
int ring_size;
--
2.11.0
^ permalink raw reply related
* [PATCH net-next 2/3] netvsc: use hv_get_bytes_to_read
From: Stephen Hemminger @ 2017-05-29 18:31 UTC (permalink / raw)
To: davem; +Cc: netdev, Stephen Hemminger
In-Reply-To: <20170529183159.22205-1-sthemmin@microsoft.com>
Don't need need to look at write space in netvsc_close.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
---
drivers/net/hyperv/netvsc_drv.c | 10 +++-------
1 file changed, 3 insertions(+), 7 deletions(-)
diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c
index 548c78863ce9..c3e0277186bb 100644
--- a/drivers/net/hyperv/netvsc_drv.c
+++ b/drivers/net/hyperv/netvsc_drv.c
@@ -118,7 +118,7 @@ static int netvsc_close(struct net_device *net)
struct net_device_context *net_device_ctx = netdev_priv(net);
struct netvsc_device *nvdev = rtnl_dereference(net_device_ctx->nvdev);
int ret;
- u32 aread, awrite, i, msec = 10, retry = 0, retry_max = 20;
+ u32 aread, i, msec = 10, retry = 0, retry_max = 20;
struct vmbus_channel *chn;
netif_tx_disable(net);
@@ -139,15 +139,11 @@ static int netvsc_close(struct net_device *net)
if (!chn)
continue;
- hv_get_ringbuffer_availbytes(&chn->inbound, &aread,
- &awrite);
-
+ aread = hv_get_bytes_to_read(&chn->inbound);
if (aread)
break;
- hv_get_ringbuffer_availbytes(&chn->outbound, &aread,
- &awrite);
-
+ aread = hv_get_bytes_to_read(&chn->outbound);
if (aread)
break;
}
--
2.11.0
^ permalink raw reply related
* [PATCH net-next 1/3] netvsc: optimize calculation of number of slots
From: Stephen Hemminger @ 2017-05-29 18:31 UTC (permalink / raw)
To: davem; +Cc: netdev, Stephen Hemminger
Speed up transmit check for fragmented packets by using existing
macros to compute number of pages, and eliminate loop since
each skb fragment is only one page.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
---
drivers/net/hyperv/netvsc_drv.c | 43 ++++++++++-------------------------------
1 file changed, 10 insertions(+), 33 deletions(-)
diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c
index 4421a6d00375..548c78863ce9 100644
--- a/drivers/net/hyperv/netvsc_drv.c
+++ b/drivers/net/hyperv/netvsc_drv.c
@@ -343,34 +343,14 @@ static u32 init_page_array(void *hdr, u32 len, struct sk_buff *skb,
return slots_used;
}
-static int count_skb_frag_slots(struct sk_buff *skb)
-{
- int i, frags = skb_shinfo(skb)->nr_frags;
- int pages = 0;
-
- for (i = 0; i < frags; i++) {
- skb_frag_t *frag = skb_shinfo(skb)->frags + i;
- unsigned long size = skb_frag_size(frag);
- unsigned long offset = frag->page_offset;
-
- /* Skip unused frames from start of page */
- offset &= ~PAGE_MASK;
- pages += PFN_UP(offset + size);
- }
- return pages;
-}
-
-static int netvsc_get_slots(struct sk_buff *skb)
+/* Estimate number of page buffers neede to transmit
+ * Need at most 2 for RNDIS header plus skb body and fragments.
+ */
+static unsigned int netvsc_get_slots(const struct sk_buff *skb)
{
- char *data = skb->data;
- unsigned int offset = offset_in_page(data);
- unsigned int len = skb_headlen(skb);
- int slots;
- int frag_slots;
-
- slots = DIV_ROUND_UP(offset + len, PAGE_SIZE);
- frag_slots = count_skb_frag_slots(skb);
- return slots + frag_slots;
+ return PFN_UP(offset_in_page(skb->data)+ skb_headlen(skb))
+ + skb_shinfo(skb)->nr_frags
+ + 2;
}
static u32 net_checksum_info(struct sk_buff *skb)
@@ -408,21 +388,18 @@ static int netvsc_start_xmit(struct sk_buff *skb, struct net_device *net)
struct hv_page_buffer page_buf[MAX_PAGE_BUFFER_COUNT];
struct hv_page_buffer *pb = page_buf;
- /* We will atmost need two pages to describe the rndis
- * header. We can only transmit MAX_PAGE_BUFFER_COUNT number
+ /* We can only transmit MAX_PAGE_BUFFER_COUNT number
* of pages in a single packet. If skb is scattered around
* more pages we try linearizing it.
*/
-
- num_data_pgs = netvsc_get_slots(skb) + 2;
-
+ num_data_pgs = netvsc_get_slots(skb);
if (unlikely(num_data_pgs > MAX_PAGE_BUFFER_COUNT)) {
++net_device_ctx->eth_stats.tx_scattered;
if (skb_linearize(skb))
goto no_memory;
- num_data_pgs = netvsc_get_slots(skb) + 2;
+ num_data_pgs = netvsc_get_slots(skb);
if (num_data_pgs > MAX_PAGE_BUFFER_COUNT) {
++net_device_ctx->eth_stats.tx_too_big;
goto drop;
--
2.11.0
^ permalink raw reply related
* [PATCH net] vxlan: eliminate cached dst leak
From: Lance Richardson @ 2017-05-29 17:25 UTC (permalink / raw)
To: netdev, pabeni
After commit 0c1d70af924b ("net: use dst_cache for vxlan device"),
cached dst entries could be leaked when more than one remote was
present for a given vxlan_fdb entry, causing subsequent netns
operations to block indefinitely and "unregister_netdevice: waiting
for lo to become free." messages to appear in the kernel log.
Fix by properly releasing cached dst and freeing resources in this
case.
Fixes: commit 0c1d70af924b ("net: use dst_cache for vxlan device")
Signed-off-by: Lance Richardson <lrichard@redhat.com>
---
drivers/net/vxlan.c | 20 +++++++++++++++++---
1 file changed, 17 insertions(+), 3 deletions(-)
diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index 328b471..5c1d69e 100644
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -740,6 +740,22 @@ static void vxlan_fdb_destroy(struct vxlan_dev *vxlan, struct vxlan_fdb *f)
call_rcu(&f->rcu, vxlan_fdb_free);
}
+static void vxlan_dst_free(struct rcu_head *head)
+{
+ struct vxlan_rdst *rd = container_of(head, struct vxlan_rdst, rcu);
+
+ dst_cache_destroy(&rd->dst_cache);
+ kfree(rd);
+}
+
+static void vxlan_fdb_dst_destroy(struct vxlan_dev *vxlan, struct vxlan_fdb *f,
+ struct vxlan_rdst *rd)
+{
+ list_del_rcu(&rd->list);
+ vxlan_fdb_notify(vxlan, f, rd, RTM_DELNEIGH);
+ call_rcu(&rd->rcu, vxlan_dst_free);
+}
+
static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan,
union vxlan_addr *ip, __be16 *port, __be32 *src_vni,
__be32 *vni, u32 *ifindex)
@@ -864,9 +880,7 @@ static int __vxlan_fdb_delete(struct vxlan_dev *vxlan,
* otherwise destroy the fdb entry
*/
if (rd && !list_is_singular(&f->remotes)) {
- list_del_rcu(&rd->list);
- vxlan_fdb_notify(vxlan, f, rd, RTM_DELNEIGH);
- kfree_rcu(rd, rcu);
+ vxlan_fdb_dst_destroy(vxlan, f, rd);
goto out;
}
--
2.7.4
^ permalink raw reply related
* RE: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
From: Ilan Tayari @ 2017-05-29 16:09 UTC (permalink / raw)
To: Ilan Tayari, Jason Gunthorpe
Cc: Alexei Starovoitov, Saeed Mahameed, David S. Miller, Doug Ledford,
netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
jsorensen-b10kYP2dOMg@public.gmane.org, Andy Shevchenko,
linux-fpga-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Alan Tull,
yi1.li-VuQAYsv1563Yd54FQh9/CA@public.gmane.org, Boris Pismenny
In-Reply-To: <AM4PR0501MB194037FF8F17466BC9ECC73DDBF30-dp/nxUn679gfNUYDR5dMTsDSnupUy6xnnBOFsp37pqbUKgpGm//BTAC/G2K4zDHf@public.gmane.org>
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="utf-8", Size: 2158 bytes --]
> -----Original Message-----
> From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org]
>
> > -----Original Message-----
> > From: Jason Gunthorpe [mailto:jgunthorpe@obsidianresearch.com]
> > Subject: Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
> >
> > On Mon, May 29, 2017 at 03:58:33PM +0000, Ilan Tayari wrote:
> > > > From: Jason Gunthorpe [mailto:jgunthorpe@obsidianresearch.com]
> > > > Subject: Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for
> > Innova
> > > >
> > > > On Sun, May 28, 2017 at 07:22:27AM +0000, Ilan Tayari wrote:
> > > >
> > > > > This is neither PCI-bar mapped, nor mailbox command.
> > > > > The FPGA is indeed a bump-on-the-wire.
> > > > > (It has I2C to the CX4 chip, but that is for debug purposes, and
> too
> > > > slow
> > > > > to perform real programming)
> > > >
> > > > Wait.. So if it truely has nothing to do with the existing mellanox
> > > > driver, then nothing more than the fpga loader should be in the mlx5
> > > > directory?
> > >
> > > True, except in specific cases when the FPGA may mangle the packets in
> > > a way that the netdevice configures, and the driver needs to adapt the
> > > data path.
> >
> > > Such is the case of IPSec and TLS offloads.
> > > Those are tied to the mlx5 Ethernet driver (isolated with a kconfig).
> >
> > But there is nothing stopping this sort of FPGA mangling logic being
> > downstream of any NIC, Mellanox is just the first to do this.
> >
> > I think you'd be better to add something to the net stack to model
> > this post-nic mangling hardware, than trying to hide it in a driver.
>
> Of course.
>
> For IPSec, this is already in the kernel.
> See this patchset:
> http://www.mail-archive.com/netdev@vger.kernel.org/msg162876.html
Sorry, I pointed at the RFC by mistake.
This is the relevant pull request:
https://patchwork.ozlabs.org/patch/752707/
>
> For TLS, Dave Watson is working with our guys to add it.
>
> >
> > Jason
N‹§²æìr¸›yúèšØb²X¬¶Ç§vØ^–)Þº{.nÇ+‰·¥Š{±ÙšŠ{ayº\x1dʇڙë,j\a¢f£¢·hš‹»öì\x17/oSc¾™Ú³9˜uÀ¦æå‰È&jw¨®\x03(階ŠÝ¢j"ú\x1a¶^[m§ÿïêäz¹Þ–Šàþf£¢·hšˆ§~ˆmš
^ permalink raw reply
* RE: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
From: Ilan Tayari @ 2017-05-29 16:05 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: Alexei Starovoitov, Saeed Mahameed, David S. Miller, Doug Ledford,
netdev@vger.kernel.org, linux-rdma@vger.kernel.org,
jsorensen@fb.com, Andy Shevchenko, linux-fpga@vger.kernel.org,
Alan Tull, yi1.li@linux.intel.com, Boris Pismenny
In-Reply-To: <20170529160252.GA27550@obsidianresearch.com>
> -----Original Message-----
> From: Jason Gunthorpe [mailto:jgunthorpe@obsidianresearch.com]
> Subject: Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for Innova
>
> On Mon, May 29, 2017 at 03:58:33PM +0000, Ilan Tayari wrote:
> > > From: Jason Gunthorpe [mailto:jgunthorpe@obsidianresearch.com]
> > > Subject: Re: [for-next 4/6] net/mlx5: FPGA, Add basic support for
> Innova
> > >
> > > On Sun, May 28, 2017 at 07:22:27AM +0000, Ilan Tayari wrote:
> > >
> > > > This is neither PCI-bar mapped, nor mailbox command.
> > > > The FPGA is indeed a bump-on-the-wire.
> > > > (It has I2C to the CX4 chip, but that is for debug purposes, and too
> > > slow
> > > > to perform real programming)
> > >
> > > Wait.. So if it truely has nothing to do with the existing mellanox
> > > driver, then nothing more than the fpga loader should be in the mlx5
> > > directory?
> >
> > True, except in specific cases when the FPGA may mangle the packets in
> > a way that the netdevice configures, and the driver needs to adapt the
> > data path.
>
> > Such is the case of IPSec and TLS offloads.
> > Those are tied to the mlx5 Ethernet driver (isolated with a kconfig).
>
> But there is nothing stopping this sort of FPGA mangling logic being
> downstream of any NIC, Mellanox is just the first to do this.
>
> I think you'd be better to add something to the net stack to model
> this post-nic mangling hardware, than trying to hide it in a driver.
Of course.
For IPSec, this is already in the kernel.
See this patchset:
http://www.mail-archive.com/netdev@vger.kernel.org/msg162876.html
For TLS, Dave Watson is working with our guys to add it.
>
> Jason
^ 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