Re: [dpdk-dev] [RFC v2 5/5] config: add WFE config entry for aarch64
>-Original Message- >From: dev On Behalf Of Gavin Hu >Sent: Wednesday, July 3, 2019 2:29 PM >To: dev@dpdk.org >Cc: n...@arm.com >Subject: [dpdk-dev] [RFC v2 5/5] config: add WFE config entry for >aarch64 > >Add the RTE_USE_WFE configuration entry for aarch64, disabled by >default. >It can be enabled selectively based on the performance benchmarking. > >Signed-off-by: Gavin Hu >Reviewed-by: Ruifeng Wang >Reviewed-by: Steve Capper >Reviewed-by: Honnappa Nagarahalli Acked-by: Pavan Nikhilesh >--- > config/arm/meson.build | 1 + > config/common_armv8a_linux | 6 ++ > 2 files changed, 7 insertions(+) >
[dpdk-dev] [PATCH 2/2] net/bnxt: fix an issue with rxq0 stop
From: Kalesh AP
For the stingray platform and other platforms needing tighter
control of resource utilization, we process async events on
receive completion ring0.
For other chips there is a dedicated completion ring for
asynchronous event handling instead of handling these events
on a receive completion ring.
This patch fixes the check for rxq0 stop.
Fixes: 05b5e4821c1e ("net/bnxt: use dedicated CPR for async events")
Reviewed-by: Lance Richardson
Signed-off-by: Kalesh AP
Signed-off-by: Ajit Khaparde
---
drivers/net/bnxt/bnxt_rxq.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c
index 4082c3aa9..8c7608621 100644
--- a/drivers/net/bnxt/bnxt_rxq.c
+++ b/drivers/net/bnxt/bnxt_rxq.c
@@ -458,8 +458,11 @@ int bnxt_rx_queue_stop(struct rte_eth_dev *dev, uint16_t
rx_queue_id)
struct bnxt_rx_queue *rxq = NULL;
int rc = 0;
- /* Rx CQ 0 also works as Default CQ for async notifications */
- if (!rx_queue_id) {
+ /* For the stingray platform and other platforms needing tighter
+* control of resource utilization, Rx CQ 0 also works as
+* Default CQ for async notifications
+*/
+ if (!BNXT_NUM_ASYNC_CPR && !rx_queue_id) {
PMD_DRV_LOG(ERR, "Cannot stop Rx queue id %d\n", rx_queue_id);
return -EINVAL;
}
--
2.20.1 (Apple Git-117)
[dpdk-dev] [PATCH 1/2] net/bnxt: fix rxq start
From: Kalesh AP
1. Added a missing return in bnxt_alloc_hwrm_rx_ring().
2. Added a missing return value check
3. Moved the log message to right place.
Fixes: 05b5e4821c1e ("net/bnxt: use dedicated CPR for async events")
Reviewed-by: Lance Richardson
Signed-off-by: Kalesh AP
Signed-off-by: Ajit Khaparde
---
drivers/net/bnxt/bnxt_ring.c | 6 ++
drivers/net/bnxt/bnxt_rxq.c | 17 +
2 files changed, 15 insertions(+), 8 deletions(-)
diff --git a/drivers/net/bnxt/bnxt_ring.c b/drivers/net/bnxt/bnxt_ring.c
index 05a9a200c..80da098ef 100644
--- a/drivers/net/bnxt/bnxt_ring.c
+++ b/drivers/net/bnxt/bnxt_ring.c
@@ -546,10 +546,8 @@ int bnxt_alloc_hwrm_rx_ring(struct bnxt *bp, int
queue_index)
bnxt_db_write(&rxr->ag_db, rxr->ag_prod);
}
rxq->index = queue_index;
- PMD_DRV_LOG(INFO,
- "queue %d, rx_deferred_start %d, state %d!\n",
- queue_index, rxq->rx_deferred_start,
- bp->eth_dev->data->rx_queue_state[queue_index]);
+
+ return 0;
err_out:
PMD_DRV_LOG(ERR,
diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c
index e0eb890f8..4082c3aa9 100644
--- a/drivers/net/bnxt/bnxt_rxq.c
+++ b/drivers/net/bnxt/bnxt_rxq.c
@@ -411,10 +411,11 @@ int bnxt_rx_queue_start(struct rte_eth_dev *dev, uint16_t
rx_queue_id)
return -EINVAL;
}
- dev->data->rx_queue_state[rx_queue_id] = RTE_ETH_QUEUE_STATE_STARTED;
-
bnxt_free_hwrm_rx_ring(bp, rx_queue_id);
- bnxt_alloc_hwrm_rx_ring(bp, rx_queue_id);
+ rc = bnxt_alloc_hwrm_rx_ring(bp, rx_queue_id);
+ if (rc)
+ return rc;
+
PMD_DRV_LOG(INFO, "Rx queue started %d\n", rx_queue_id);
if (dev_conf->rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG) {
@@ -435,8 +436,16 @@ int bnxt_rx_queue_start(struct rte_eth_dev *dev, uint16_t
rx_queue_id)
rc = bnxt_vnic_rss_configure(bp, vnic);
}
- if (rc == 0)
+ if (rc == 0) {
+ dev->data->rx_queue_state[rx_queue_id] =
+ RTE_ETH_QUEUE_STATE_STARTED;
rxq->rx_deferred_start = false;
+ }
+
+ PMD_DRV_LOG(INFO,
+ "queue %d, rx_deferred_start %d, state %d!\n",
+ rx_queue_id, rxq->rx_deferred_start,
+ bp->eth_dev->data->rx_queue_state[rx_queue_id]);
return rc;
}
--
2.20.1 (Apple Git-117)
[dpdk-dev] [v2] net/e1000: fix buffer overrun while i219 processing DMA transactions
Intel® 100/200 Series Chipset platforms reduced the round-trip
latency for the LAN Controller DMA accesses, causing in some high
performance cases a buffer overrun while the I219 LAN Connected
Device is processing the DMA transactions. I219LM and I219V devices
can fall into unrecovered Tx hang under very stressfully UDP traffic
and multiple reconnection of Ethernet cable. This Tx hang of the LAN
Controller is only recovered if the system is rebooted. Slightly slow
down DMA access by reducing the number of outstanding requests.
This workaround could have an impact on TCP traffic performance
on the platform. Disabling TSO eliminates performance loss for TCP
traffic without a noticeable impact on CPU performance.
Please, refer to I218/I219 specification update:
https://www.intel.com/content/www/us/en/embedded/products/networking/
ethernet-connection-i218-family-documentation.html
Cc: sta...@dpdk.org
Signed-off-by: Xiao Zhang
---
v2
Change the fix from igb_rxtx to em_rxtx, since i219 use em PMD.
v1
Apply kernel fix to igb PMD driver.
---
drivers/net/e1000/base/e1000_ich8lan.h | 1 +
drivers/net/e1000/em_rxtx.c| 16
2 files changed, 17 insertions(+)
diff --git a/drivers/net/e1000/base/e1000_ich8lan.h
b/drivers/net/e1000/base/e1000_ich8lan.h
index bc4ed1d..9ee94f6 100644
--- a/drivers/net/e1000/base/e1000_ich8lan.h
+++ b/drivers/net/e1000/base/e1000_ich8lan.h
@@ -133,6 +133,7 @@ POSSIBILITY OF SUCH DAMAGE.
#define E1000_FLASH_BASE_ADDR 0xE000 /*offset of NVM access regs*/
#define E1000_CTRL_EXT_NVMVS 0x3 /*NVM valid sector */
#define E1000_TARC0_CB_MULTIQ_3_REQ(1 << 28 | 1 << 29)
+#define E1000_TARC0_CB_MULTIQ_2_REQ(1 << 29)
#define PCIE_ICH8_SNOOP_ALLPCIE_NO_SNOOP_ALL
#define E1000_ICH_RAR_ENTRIES 7
diff --git a/drivers/net/e1000/em_rxtx.c b/drivers/net/e1000/em_rxtx.c
index 0d0d38e..d53ee80 100644
--- a/drivers/net/e1000/em_rxtx.c
+++ b/drivers/net/e1000/em_rxtx.c
@@ -1970,6 +1970,22 @@ eth_em_tx_init(struct rte_eth_dev *dev)
tctl |= (E1000_TCTL_PSP | E1000_TCTL_RTLC | E1000_TCTL_EN |
(E1000_COLLISION_THRESHOLD << E1000_CT_SHIFT));
+ /* SPT and CNP Si errata workaround to avoid data corruption */
+ if (hw->mac.type == e1000_pch_spt) {
+ uint32_t reg_val;
+ reg_val = E1000_READ_REG(hw, E1000_IOSFPC);
+ reg_val |= E1000_RCTL_RDMTS_HEX;
+ E1000_WRITE_REG(hw, E1000_IOSFPC, reg_val);
+
+ /* Dropping the number of outstanding requests from
+* 3 to 2 in order to avoid a buffer overrun.
+*/
+ reg_val = E1000_READ_REG(hw, E1000_TARC(0));
+ reg_val &= ~E1000_TARC0_CB_MULTIQ_3_REQ;
+ reg_val |= E1000_TARC0_CB_MULTIQ_2_REQ;
+ E1000_WRITE_REG(hw, E1000_TARC(0), reg_val);
+ }
+
/* This write will effectively turn on the transmit unit. */
E1000_WRITE_REG(hw, E1000_TCTL, tctl);
}
--
2.7.4
Re: [dpdk-dev] [DPDK] net/e1000: fix buffer overrun while i219 processing DMA transactions
> -Original Message-
> From: Anand H. Krishnan [mailto:anandhkrish...@gmail.com]
> Sent: Saturday, July 20, 2019 10:57 AM
> To: Zhang, Xiao
> Cc: dev@dpdk.org; Lu, Wenzhuo ; Zhao1, Wei
>
> Subject: Re: [dpdk-dev] [DPDK] net/e1000: fix buffer overrun while i219
> processing DMA transactions
>
> This seems to be changing the IGB driver. Shouldn't you be changing the em
> driver rather than the igb driver?
Yes, the fix changed to em driver in v2 patch. Thanks.
> Thanks,
> Anand
>
> On Mon, Jul 8, 2019 at 10:10 AM Xiao Zhang wrote:
> >
> > Intel® 100/200 Series Chipset platforms reduced the round-trip latency
> > for the LAN Controller DMA accesses, causing in some high performance
> > cases a buffer overrun while the I219 LAN Connected Device is
> > processing the DMA transactions. I219LM and I219V devices can fall
> > into unrecovered Tx hang under very stressfully UDP traffic and
> > multiple reconnection of Ethernet cable. This Tx hang of the LAN
> > Controller is only recovered if the system is rebooted. Slightly slow
> > down DMA access by reducing the number of outstanding requests.
> > This workaround could have an impact on TCP traffic performance on the
> > platform. Disabling TSO eliminates performance loss for TCP traffic
> > without a noticeable impact on CPU performance.
> >
> > Please, refer to I218/I219 specification update:
> >
> https://www.intel.com/content/www/us/en/embedded/products/network
> ing/
> > ethernet-connection-i218-family-documentation.html
> >
> > Signed-off-by: Xiao Zhang
> > ---
> > drivers/net/e1000/base/e1000_ich8lan.h | 1 +
> > drivers/net/e1000/igb_rxtx.c | 16
> > 2 files changed, 17 insertions(+)
> >
> > diff --git a/drivers/net/e1000/base/e1000_ich8lan.h
> > b/drivers/net/e1000/base/e1000_ich8lan.h
> > index 1f2a3f8..084eb9c 100644
> > --- a/drivers/net/e1000/base/e1000_ich8lan.h
> > +++ b/drivers/net/e1000/base/e1000_ich8lan.h
> > @@ -134,6 +134,7 @@ POSSIBILITY OF SUCH DAMAGE.
> > #define E1000_FLASH_BASE_ADDR 0xE000 /*offset of NVM access regs*/
> > #define E1000_CTRL_EXT_NVMVS 0x3 /*NVM valid sector */
> > #define E1000_TARC0_CB_MULTIQ_3_REQ(1 << 28 | 1 << 29)
> > +#define E1000_TARC0_CB_MULTIQ_2_REQ(1 << 29)
> > #define PCIE_ICH8_SNOOP_ALLPCIE_NO_SNOOP_ALL
> >
> > #define E1000_ICH_RAR_ENTRIES 7
> > diff --git a/drivers/net/e1000/igb_rxtx.c
> > b/drivers/net/e1000/igb_rxtx.c index 33eeb4e..5d45e62 100644
> > --- a/drivers/net/e1000/igb_rxtx.c
> > +++ b/drivers/net/e1000/igb_rxtx.c
> > @@ -2627,6 +2627,22 @@ eth_igb_tx_init(struct rte_eth_dev *dev)
> >
> > e1000_config_collision_dist(hw);
> >
> > + /* SPT and CNP Si errata workaround to avoid data corruption */
> > + if (hw->mac.type == e1000_pch_spt) {
> > + uint32_t reg_val;
> > + reg_val = E1000_READ_REG(hw, E1000_IOSFPC);
> > + reg_val |= E1000_RCTL_RDMTS_HEX;
> > + E1000_WRITE_REG(hw, E1000_IOSFPC, reg_val);
> > +
> > + /* Dropping the number of outstanding requests from
> > +* 3 to 2 in order to avoid a buffer overrun.
> > +*/
> > + reg_val = E1000_READ_REG(hw, E1000_TARC(0));
> > + reg_val &= ~E1000_TARC0_CB_MULTIQ_3_REQ;
> > + reg_val |= E1000_TARC0_CB_MULTIQ_2_REQ;
> > + E1000_WRITE_REG(hw, E1000_TARC(0), reg_val);
> > + }
> > +
> > /* This write will effectively turn on the transmit unit. */
> > E1000_WRITE_REG(hw, E1000_TCTL, tctl); }
> > --
> > 2.7.4
> >
