RE: [RFC] eal: use same atomic intrinsics for gcc and clang
> From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
> Sent: Saturday, 11 February 2023 02.56
>
> The size generic atomic intrinsics generate the same
> code as the size specific intrinsics for gcc. Use size
> generic intrinsics for both gcc and clang.
>
> Fixes: 7bdccb93078e ("eal: fix ARM build with clang")
> Cc: sta...@dpdk.org
> Cc: pbhagavat...@marvell.com
>
> Signed-off-by: Honnappa Nagarahalli
> ---
Acked-by: Morten Brørup
Re: [dpdk-dev] [PATCH v2 3/3] common/cnxk: fix inline IPsec rule creation
On Fri, Feb 10, 2023 at 1:56 PM wrote:
>
> From: Satheesh Paul
>
> Use inline IPsec device to create IPsec rules irrespective
> of RTE_ETH_RX_OFFLOAD_SECURITY offload flag.
>
> Fixes: 1aa86a170e27 ("drivers: support IPsec rule reservation scheme")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Satheesh Paul
> Reviewed-by: Nithin Dabilpuram
> Reviewed-by: Kiran Kumar K
Series applied to dpdk-next-net-eventdev/for-main. Thanks
> ---
> drivers/common/cnxk/roc_npc.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/common/cnxk/roc_npc.c b/drivers/common/cnxk/roc_npc.c
> index 579b3dce9a..d3efebe663 100644
> --- a/drivers/common/cnxk/roc_npc.c
> +++ b/drivers/common/cnxk/roc_npc.c
> @@ -562,8 +562,10 @@ npc_parse_actions(struct roc_npc *roc_npc, const struct
> roc_npc_attr *attr,
> }
> rq = inl_rq->qid;
> pf_func = nix_inl_dev_pffunc_get();
> - flow->is_inline_dev = 1;
> }
> +
> + if (roc_nix_inl_dev_is_probed())
> + flow->is_inline_dev = 1;
> sec_action = actions;
> break;
> case ROC_NPC_ACTION_TYPE_VLAN_STRIP:
> --
> 2.35.3
>
[PATCH 0/6] net/hns3: some code optimization for Rx/Tx function
This patchset is to do some code optimization for Rx/Tx function. Huisong Li (6): net/hns3: fix querying dummy function failed net/hns3: add debug info for Rx/Tx dummy function net/hns3: remove debug condition for Tx prepare net/hns3: separate tx prepare from getting Tx function net/hns3: remove getting Tx function from head file net/hns3: extract common functions to set Rx/Tx drivers/net/hns3/hns3_ethdev.c| 20 ++ drivers/net/hns3/hns3_ethdev_vf.c | 19 ++ drivers/net/hns3/hns3_mp.c| 4 +- drivers/net/hns3/hns3_rxtx.c | 108 ++ drivers/net/hns3/hns3_rxtx.h | 5 +- 5 files changed, 79 insertions(+), 77 deletions(-) -- 2.22.0
[PATCH 1/6] net/hns3: fix querying dummy function failed
From: Huisong Li
The rte_eth_rx/tx_burst_mode_get API will fail when Rx/Tx
function is dummy.
Fixes: 7ef933908f04 ("net/hns3: add simple Tx path")
Fixes: 521ab3e93361 ("net/hns3: add simple Rx path")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_rxtx.c | 38 ++--
1 file changed, 23 insertions(+), 15 deletions(-)
diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index f1163ce8a9..feef9e1c7b 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -2786,6 +2786,7 @@ hns3_rx_burst_mode_get(struct rte_eth_dev *dev,
__rte_unused uint16_t queue_id,
{ hns3_recv_scattered_pkts, "Scalar Scattered" },
{ hns3_recv_pkts_vec, "Vector Neon" },
{ hns3_recv_pkts_vec_sve, "Vector Sve"},
+ { rte_eth_pkt_burst_dummy, "Dummy" },
};
eth_rx_burst_t pkt_burst = dev->rx_pkt_burst;
@@ -4272,24 +4273,31 @@ int
hns3_tx_burst_mode_get(struct rte_eth_dev *dev, __rte_unused uint16_t queue_id,
struct rte_eth_burst_mode *mode)
{
+ static const struct {
+ eth_tx_burst_t pkt_burst;
+ const char *info;
+ } burst_infos[] = {
+ { hns3_xmit_pkts_simple,"Scalar Simple" },
+ { hns3_xmit_pkts, "Scalar"},
+ { hns3_xmit_pkts_vec, "Vector Neon" },
+ { hns3_xmit_pkts_vec_sve, "Vector Sve"},
+ { rte_eth_pkt_burst_dummy, "Dummy" },
+ };
+
eth_tx_burst_t pkt_burst = dev->tx_pkt_burst;
- const char *info = NULL;
-
- if (pkt_burst == hns3_xmit_pkts_simple)
- info = "Scalar Simple";
- else if (pkt_burst == hns3_xmit_pkts)
- info = "Scalar";
- else if (pkt_burst == hns3_xmit_pkts_vec)
- info = "Vector Neon";
- else if (pkt_burst == hns3_xmit_pkts_vec_sve)
- info = "Vector Sve";
-
- if (info == NULL)
- return -EINVAL;
+ int ret = -EINVAL;
+ unsigned int i;
- snprintf(mode->info, sizeof(mode->info), "%s", info);
+ for (i = 0; i < RTE_DIM(burst_infos); i++) {
+ if (pkt_burst == burst_infos[i].pkt_burst) {
+ snprintf(mode->info, sizeof(mode->info), "%s",
+burst_infos[i].info);
+ ret = 0;
+ break;
+ }
+ }
- return 0;
+ return ret;
}
static bool
--
2.22.0
[PATCH 2/6] net/hns3: add debug info for Rx/Tx dummy function
From: Huisong Li
Now dummy function can be report by rte_eth_rx/tx_burst_mode_get.
So this patch adds debug info for Rx/Tx dummy function.
Fixes: 7feb2aee0e2c ("net/hns3: log selected datapath")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index feef9e1c7b..2524f4c3c2 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -4420,13 +4420,13 @@ hns3_set_rxtx_function(struct rte_eth_dev *eth_dev)
hns3_get_tx_function(eth_dev, &prep);
eth_dev->tx_pkt_prepare = prep;
eth_dev->tx_descriptor_status = hns3_dev_tx_descriptor_status;
- hns3_trace_rxtx_function(eth_dev);
} else {
eth_dev->rx_pkt_burst = rte_eth_pkt_burst_dummy;
eth_dev->tx_pkt_burst = rte_eth_pkt_burst_dummy;
eth_dev->tx_pkt_prepare = NULL;
}
+ hns3_trace_rxtx_function(eth_dev);
hns3_eth_dev_fp_ops_config(eth_dev);
}
--
2.22.0
[PATCH 3/6] net/hns3: remove debug condition for Tx prepare
From: Huisong Li
The Tx prepare in driver is always needed if RTE_LIBRTE_ETHDEV_DEBUG
is defined. But it doesn't matter with this macro. Let's remove it.
Fixes: d7ec2c076579 ("net/hns3: select Tx prepare based on Tx offload")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_rxtx.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index 2524f4c3c2..921d72d69d 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -4311,11 +4311,6 @@ hns3_tx_check_simple_support(struct rte_eth_dev *dev)
static bool
hns3_get_tx_prep_needed(struct rte_eth_dev *dev)
{
-#ifdef RTE_LIBRTE_ETHDEV_DEBUG
- RTE_SET_USED(dev);
- /* always perform tx_prepare when debug */
- return true;
-#else
#define HNS3_DEV_TX_CSKUM_TSO_OFFLOAD_MASK (\
RTE_ETH_TX_OFFLOAD_IPV4_CKSUM | \
RTE_ETH_TX_OFFLOAD_TCP_CKSUM | \
@@ -4333,7 +4328,6 @@ hns3_get_tx_prep_needed(struct rte_eth_dev *dev)
return true;
return false;
-#endif
}
eth_tx_burst_t
--
2.22.0
[PATCH 4/6] net/hns3: separate tx prepare from getting Tx function
From: Huisong Li
Separate getting tx prepare from hns3_get_tx_function by extracting
an independent function.
Fixes: d7ec2c076579 ("net/hns3: select Tx prepare based on Tx offload")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_rxtx.c | 32 ++--
drivers/net/hns3/hns3_rxtx.h | 3 +--
2 files changed, 15 insertions(+), 20 deletions(-)
diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index 921d72d69d..9b2bd02857 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -4324,26 +4324,30 @@ hns3_get_tx_prep_needed(struct rte_eth_dev *dev)
RTE_ETH_TX_OFFLOAD_GENEVE_TNL_TSO)
uint64_t tx_offload = dev->data->dev_conf.txmode.offloads;
+
if (tx_offload & HNS3_DEV_TX_CSKUM_TSO_OFFLOAD_MASK)
return true;
return false;
}
+static eth_tx_prep_t
+hns3_get_tx_prepare(struct rte_eth_dev *dev)
+{
+ return hns3_get_tx_prep_needed(dev) ? hns3_prep_pkts : NULL;
+}
+
eth_tx_burst_t
-hns3_get_tx_function(struct rte_eth_dev *dev, eth_tx_prep_t *prep)
+hns3_get_tx_function(struct rte_eth_dev *dev)
{
struct hns3_adapter *hns = dev->data->dev_private;
bool vec_allowed, sve_allowed, simple_allowed;
- bool vec_support, tx_prepare_needed;
+ bool vec_support;
vec_support = hns3_tx_check_vec_support(dev) == 0;
vec_allowed = vec_support && hns3_get_default_vec_support();
sve_allowed = vec_support && hns3_get_sve_support();
simple_allowed = hns3_tx_check_simple_support(dev);
- tx_prepare_needed = hns3_get_tx_prep_needed(dev);
-
- *prep = NULL;
if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_VEC && vec_allowed)
return hns3_xmit_pkts_vec;
@@ -4351,19 +4355,14 @@ hns3_get_tx_function(struct rte_eth_dev *dev,
eth_tx_prep_t *prep)
return hns3_xmit_pkts_vec_sve;
if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_SIMPLE && simple_allowed)
return hns3_xmit_pkts_simple;
- if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_COMMON) {
- if (tx_prepare_needed)
- *prep = hns3_prep_pkts;
+ if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_COMMON)
return hns3_xmit_pkts;
- }
if (vec_allowed)
return hns3_xmit_pkts_vec;
if (simple_allowed)
return hns3_xmit_pkts_simple;
- if (tx_prepare_needed)
- *prep = hns3_prep_pkts;
return hns3_xmit_pkts;
}
@@ -4403,7 +4402,6 @@ hns3_set_rxtx_function(struct rte_eth_dev *eth_dev)
{
struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(eth_dev->data->dev_private);
struct hns3_adapter *hns = eth_dev->data->dev_private;
- eth_tx_prep_t prep = NULL;
if (hns->hw.adapter_state == HNS3_NIC_STARTED &&
__atomic_load_n(&hns->hw.reset.resetting, __ATOMIC_RELAXED) == 0) {
@@ -4411,8 +4409,8 @@ hns3_set_rxtx_function(struct rte_eth_dev *eth_dev)
eth_dev->rx_descriptor_status = hns3_dev_rx_descriptor_status;
eth_dev->tx_pkt_burst = hw->set_link_down ?
rte_eth_pkt_burst_dummy :
- hns3_get_tx_function(eth_dev, &prep);
- eth_dev->tx_pkt_prepare = prep;
+ hns3_get_tx_function(eth_dev);
+ eth_dev->tx_pkt_prepare = hns3_get_tx_prepare(eth_dev);
eth_dev->tx_descriptor_status = hns3_dev_tx_descriptor_status;
} else {
eth_dev->rx_pkt_burst = rte_eth_pkt_burst_dummy;
@@ -4758,10 +4756,8 @@ hns3_stop_tx_datapath(struct rte_eth_dev *dev)
void
hns3_start_tx_datapath(struct rte_eth_dev *dev)
{
- eth_tx_prep_t prep = NULL;
-
- dev->tx_pkt_burst = hns3_get_tx_function(dev, &prep);
- dev->tx_pkt_prepare = prep;
+ dev->tx_pkt_burst = hns3_get_tx_function(dev);
+ dev->tx_pkt_prepare = hns3_get_tx_prepare(dev);
hns3_eth_dev_fp_ops_config(dev);
if (rte_eal_process_type() == RTE_PROC_SECONDARY)
diff --git a/drivers/net/hns3/hns3_rxtx.h b/drivers/net/hns3/hns3_rxtx.h
index ea1a805491..38c3581312 100644
--- a/drivers/net/hns3/hns3_rxtx.h
+++ b/drivers/net/hns3/hns3_rxtx.h
@@ -740,8 +740,7 @@ int hns3_tx_burst_mode_get(struct rte_eth_dev *dev,
const uint32_t *hns3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
void hns3_init_rx_ptype_tble(struct rte_eth_dev *dev);
void hns3_set_rxtx_function(struct rte_eth_dev *eth_dev);
-eth_tx_burst_t hns3_get_tx_function(struct rte_eth_dev *dev,
- eth_tx_prep_t *prep);
+eth_tx_burst_t hns3_get_tx_function(struct rte_eth_dev *dev);
uint32_t hns3_get_tqp_intr_reg_offset(uint16_t tqp_intr_id);
void hns3_set_queue_intr_gl(struct hns3_hw *hw, uint16_t queue_id,
--
2.22.0
[PATCH 5/6] net/hns3: remove getting Tx function from head file
From: Huisong Li
The hns3_get_tx_function() is an intrinsic function now and should
not be open to other files.
Fixes: 96c33cfb06cf ("net/hns3: fix Rx/Tx functions update")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_rxtx.c | 2 +-
drivers/net/hns3/hns3_rxtx.h | 2 --
2 files changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index 9b2bd02857..def22bcd30 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -4337,7 +4337,7 @@ hns3_get_tx_prepare(struct rte_eth_dev *dev)
return hns3_get_tx_prep_needed(dev) ? hns3_prep_pkts : NULL;
}
-eth_tx_burst_t
+static eth_tx_burst_t
hns3_get_tx_function(struct rte_eth_dev *dev)
{
struct hns3_adapter *hns = dev->data->dev_private;
diff --git a/drivers/net/hns3/hns3_rxtx.h b/drivers/net/hns3/hns3_rxtx.h
index 38c3581312..1bdc124b7b 100644
--- a/drivers/net/hns3/hns3_rxtx.h
+++ b/drivers/net/hns3/hns3_rxtx.h
@@ -740,8 +740,6 @@ int hns3_tx_burst_mode_get(struct rte_eth_dev *dev,
const uint32_t *hns3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
void hns3_init_rx_ptype_tble(struct rte_eth_dev *dev);
void hns3_set_rxtx_function(struct rte_eth_dev *eth_dev);
-eth_tx_burst_t hns3_get_tx_function(struct rte_eth_dev *dev);
-
uint32_t hns3_get_tqp_intr_reg_offset(uint16_t tqp_intr_id);
void hns3_set_queue_intr_gl(struct hns3_hw *hw, uint16_t queue_id,
uint8_t gl_idx, uint16_t gl_value);
--
2.22.0
[PATCH 6/6] net/hns3: extract common functions to set Rx/Tx
From: Huisong Li
Extract two common functions to set Rx/Tx function in order to
reduce duplicate codes.
Fixes: 23d4b61fee5d ("net/hns3: support multiple process")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_ethdev.c| 20
drivers/net/hns3/hns3_ethdev_vf.c | 19 ---
drivers/net/hns3/hns3_mp.c| 4 ++--
drivers/net/hns3/hns3_rxtx.c | 28
drivers/net/hns3/hns3_rxtx.h | 2 ++
5 files changed, 40 insertions(+), 33 deletions(-)
diff --git a/drivers/net/hns3/hns3_ethdev.c b/drivers/net/hns3/hns3_ethdev.c
index ab565ce128..6babf67fce 100644
--- a/drivers/net/hns3/hns3_ethdev.c
+++ b/drivers/net/hns3/hns3_ethdev.c
@@ -5052,8 +5052,7 @@ hns3_dev_start(struct rte_eth_dev *dev)
rte_spinlock_unlock(&hw->lock);
hns3_rx_scattered_calc(dev);
- hns3_set_rxtx_function(dev);
- hns3_mp_req_start_rxtx(dev);
+ hns3_start_rxtx_datapath(dev);
/* Enable interrupt of all rx queues before enabling queues */
hns3_dev_all_rx_queue_intr_enable(hw, true);
@@ -5131,12 +5130,7 @@ hns3_dev_stop(struct rte_eth_dev *dev)
dev->data->dev_started = 0;
hw->adapter_state = HNS3_NIC_STOPPING;
- hns3_set_rxtx_function(dev);
- rte_wmb();
- /* Disable datapath on secondary process. */
- hns3_mp_req_stop_rxtx(dev);
- /* Prevent crashes when queues are still in use. */
- rte_delay_ms(hw->cfg_max_queues);
+ hns3_stop_rxtx_datapath(dev);
rte_spinlock_lock(&hw->lock);
if (__atomic_load_n(&hw->reset.resetting, __ATOMIC_RELAXED) == 0) {
@@ -5752,12 +5746,7 @@ hns3_stop_service(struct hns3_adapter *hns)
rte_eal_alarm_cancel(hns3_service_handler, eth_dev);
hns3_update_linkstatus_and_event(hw, false);
}
-
- hns3_set_rxtx_function(eth_dev);
- rte_wmb();
- /* Disable datapath on secondary process. */
- hns3_mp_req_stop_rxtx(eth_dev);
- rte_delay_ms(hw->cfg_max_queues);
+ hns3_stop_rxtx_datapath(eth_dev);
rte_spinlock_lock(&hw->lock);
if (hns->hw.adapter_state == HNS3_NIC_STARTED ||
@@ -5790,8 +5779,7 @@ hns3_start_service(struct hns3_adapter *hns)
hw->reset.level == HNS3_GLOBAL_RESET)
hns3_set_rst_done(hw);
eth_dev = &rte_eth_devices[hw->data->port_id];
- hns3_set_rxtx_function(eth_dev);
- hns3_mp_req_start_rxtx(eth_dev);
+ hns3_start_rxtx_datapath(eth_dev);
if (hw->adapter_state == HNS3_NIC_STARTED) {
/*
* This API parent function already hold the hns3_hw.lock, the
diff --git a/drivers/net/hns3/hns3_ethdev_vf.c
b/drivers/net/hns3/hns3_ethdev_vf.c
index d3c1bdf2c5..d051a1357b 100644
--- a/drivers/net/hns3/hns3_ethdev_vf.c
+++ b/drivers/net/hns3/hns3_ethdev_vf.c
@@ -1572,12 +1572,7 @@ hns3vf_dev_stop(struct rte_eth_dev *dev)
dev->data->dev_started = 0;
hw->adapter_state = HNS3_NIC_STOPPING;
- hns3_set_rxtx_function(dev);
- rte_wmb();
- /* Disable datapath on secondary process. */
- hns3_mp_req_stop_rxtx(dev);
- /* Prevent crashes when queues are still in use. */
- rte_delay_ms(hw->cfg_max_queues);
+ hns3_stop_rxtx_datapath(dev);
rte_spinlock_lock(&hw->lock);
if (__atomic_load_n(&hw->reset.resetting, __ATOMIC_RELAXED) == 0) {
@@ -1731,8 +1726,7 @@ hns3vf_dev_start(struct rte_eth_dev *dev)
rte_spinlock_unlock(&hw->lock);
hns3_rx_scattered_calc(dev);
- hns3_set_rxtx_function(dev);
- hns3_mp_req_start_rxtx(dev);
+ hns3_start_rxtx_datapath(dev);
/* Enable interrupt of all rx queues before enabling queues */
hns3_dev_all_rx_queue_intr_enable(hw, true);
@@ -1902,11 +1896,7 @@ hns3vf_stop_service(struct hns3_adapter *hns)
}
hw->mac.link_status = RTE_ETH_LINK_DOWN;
- hns3_set_rxtx_function(eth_dev);
- rte_wmb();
- /* Disable datapath on secondary process. */
- hns3_mp_req_stop_rxtx(eth_dev);
- rte_delay_ms(hw->cfg_max_queues);
+ hns3_stop_rxtx_datapath(eth_dev);
rte_spinlock_lock(&hw->lock);
if (hw->adapter_state == HNS3_NIC_STARTED ||
@@ -1938,8 +1928,7 @@ hns3vf_start_service(struct hns3_adapter *hns)
struct rte_eth_dev *eth_dev;
eth_dev = &rte_eth_devices[hw->data->port_id];
- hns3_set_rxtx_function(eth_dev);
- hns3_mp_req_start_rxtx(eth_dev);
+ hns3_start_rxtx_datapath(eth_dev);
rte_eal_alarm_set(HNS3VF_KEEP_ALIVE_INTERVAL, hns3vf_keep_alive_handler,
eth_dev);
diff --git a/drivers/net/hns3/hns3_mp.c b/drivers/net/hns3/hns3_mp.c
index 7184f9ad58..556f1941c6 100644
--- a/drivers/net/hns3/hns3_mp.c
+++ b/drivers/net/hns3/hns3_mp.c
@@ -89,12 +89,12 @@ mp_secondary_handle(const struct rte_mp_msg *mp_msg, const
void *peer)
case H
Re: [PATCH v2] common/cnxk: remove unnecessary locks
On Wed, Feb 8, 2023 at 1:58 AM wrote:
>
> From: Pavan Nikhilesh
>
> Remove unnecessary locks as locking is now taken care by
> mbox_get and mbox_put.
>
> Signed-off-by: Pavan Nikhilesh
Updated the git commit as follows and applied to
dpdk-next-net-mrvl/for-next-net. Thanks
common/cnxk: remove redundant locks
Remove redundant locks as locking is now taken care by
mbox_get() and mbox_put().
Signed-off-by: Pavan Nikhilesh
> ---
> v2 Changes:
> - Rebase on main
>
> drivers/common/cnxk/roc_sso.c | 25 -
> drivers/common/cnxk/roc_sso_priv.h | 1 -
> 2 files changed, 26 deletions(-)
>
> diff --git a/drivers/common/cnxk/roc_sso.c b/drivers/common/cnxk/roc_sso.c
> index a422e37be6..9e466f09ef 100644
> --- a/drivers/common/cnxk/roc_sso.c
> +++ b/drivers/common/cnxk/roc_sso.c
> @@ -327,7 +327,6 @@ roc_sso_hws_stats_get(struct roc_sso *roc_sso, uint8_t
> hws,
> struct mbox *mbox;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> mbox = mbox_get(dev->mbox);
> req_rsp = (struct sso_hws_stats *)mbox_alloc_msg_sso_hws_get_stats(
> mbox);
> @@ -354,7 +353,6 @@ roc_sso_hws_stats_get(struct roc_sso *roc_sso, uint8_t
> hws,
> stats->arbitration = req_rsp->arbitration;
> fail:
> mbox_put(mbox);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -368,7 +366,6 @@ roc_sso_hwgrp_stats_get(struct roc_sso *roc_sso, uint8_t
> hwgrp,
> struct mbox *mbox;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> mbox = mbox_get(dev->mbox);
> req_rsp = (struct sso_grp_stats *)mbox_alloc_msg_sso_grp_get_stats(
> mbox);
> @@ -403,7 +400,6 @@ roc_sso_hwgrp_stats_get(struct roc_sso *roc_sso, uint8_t
> hwgrp,
>
> fail:
> mbox_put(mbox);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -427,7 +423,6 @@ roc_sso_hwgrp_qos_config(struct roc_sso *roc_sso, struct
> roc_sso_hwgrp_qos *qos,
> struct mbox *mbox;
> int i, rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> mbox = mbox_get(dev->mbox);
> for (i = 0; i < nb_qos; i++) {
> uint8_t iaq_prcnt = qos[i].iaq_prcnt;
> @@ -461,7 +456,6 @@ roc_sso_hwgrp_qos_config(struct roc_sso *roc_sso, struct
> roc_sso_hwgrp_qos *qos,
> rc = -EIO;
> fail:
> mbox_put(mbox);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -562,11 +556,9 @@ roc_sso_hwgrp_init_xaq_aura(struct roc_sso *roc_sso,
> uint32_t nb_xae)
> struct dev *dev = &sso->dev;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> rc = sso_hwgrp_init_xaq_aura(dev, &roc_sso->xaq, nb_xae,
> roc_sso->xae_waes, roc_sso->xaq_buf_size,
> roc_sso->nb_hwgrp);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -600,9 +592,7 @@ roc_sso_hwgrp_free_xaq_aura(struct roc_sso *roc_sso,
> uint16_t nb_hwgrp)
> struct dev *dev = &sso->dev;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> rc = sso_hwgrp_free_xaq_aura(dev, &roc_sso->xaq, nb_hwgrp);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -639,9 +629,7 @@ roc_sso_hwgrp_alloc_xaq(struct roc_sso *roc_sso, uint32_t
> npa_aura_id,
> struct dev *dev = &sso->dev;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> rc = sso_hwgrp_alloc_xaq(dev, npa_aura_id, hwgrps);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -677,9 +665,7 @@ roc_sso_hwgrp_release_xaq(struct roc_sso *roc_sso,
> uint16_t hwgrps)
> struct dev *dev = &sso->dev;
> int rc;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> rc = sso_hwgrp_release_xaq(dev, hwgrps);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -693,7 +679,6 @@ roc_sso_hwgrp_set_priority(struct roc_sso *roc_sso,
> uint16_t hwgrp,
> struct mbox *mbox;
> int rc = -ENOSPC;
>
> - plt_spinlock_lock(&sso->mbox_lock);
> mbox = mbox_get(dev->mbox);
> req = mbox_alloc_msg_sso_grp_set_priority(mbox);
> if (req == NULL)
> @@ -709,14 +694,12 @@ roc_sso_hwgrp_set_priority(struct roc_sso *roc_sso,
> uint16_t hwgrp,
> goto fail;
> }
> mbox_put(mbox);
> - plt_spinlock_unlock(&sso->mbox_lock);
> plt_sso_dbg("HWGRP %d weight %d affinity %d priority %d", hwgrp,
> weight,
> affinity, priority);
>
> return 0;
> fail:
> mbox_put(mbox);
> - plt_spinlock_unlock(&sso->mbox_lock);
> return rc;
> }
>
> @@ -778,7 +761,6 @@ roc_sso_rsrc_init(struct roc_sso *roc_sso, uint8_t
> nb_hws, uint16_t nb_hwgrp, ui
> if (!nb_hws || roc_sso->max_hws < nb_hws)
> return -ENOENT;
>
[Bug 1159] lib/vhost build failure with clang 3.4.2
https://bugs.dpdk.org/show_bug.cgi?id=1159 Bug ID: 1159 Summary: lib/vhost build failure with clang 3.4.2 Product: DPDK Version: 23.03 Hardware: All OS: All Status: UNCONFIRMED Severity: normal Priority: Normal Component: vhost/virtio Assignee: dev@dpdk.org Reporter: alia...@nvidia.com Target Milestone: --- ``` FAILED: lib/librte_vhost.a.p/vhost_iotlb.c.o clang -Ilib/librte_vhost.a.p -Ilib -I../../root/dpdk/lib -Ilib/vhost -I../../root/dpdk/lib/vhost -I. -I../../root/dpdk -Iconfig -I../../root/dpdk/config -Ilib/eal/include -I../../root/dpdk/lib/eal/include -Ilib/eal/linux/include -I../../root/dpdk/lib/eal/linux/include -Ilib/eal/x86/include -I../../root/dpdk/lib/eal/x86/include -Ilib/eal/common -I../../root/dpdk/lib/eal/common -Ilib/eal -I../../root/dpdk/lib/eal -Ilib/kvargs -I../../root/dpdk/lib/kvargs -Ilib/metrics -I../../root/dpdk/lib/metrics -Ilib/telemetry -I../../root/dpdk/lib/telemetry -Ilib/ethdev -I../../root/dpdk/lib/ethdev -Ilib/net -I../../root/dpdk/lib/net -Ilib/mbuf -I../../root/dpdk/lib/mbuf -Ilib/mempool -I../../root/dpdk/lib/mempool -Ilib/ring -I../../root/dpdk/lib/ring -Ilib/meter -I../../root/dpdk/lib/meter -Ilib/cryptodev -I../../root/dpdk/lib/cryptodev -Ilib/rcu -I../../root/dpdk/lib/rcu -Ilib/hash -I../../root/dpdk/lib/hash -Ilib/pci -I../../root/dpdk/lib/pci -Ilib/dmadev -I../../root/dpdk/lib/dmadev -Xclang -fcolor-diagnostics -pipe -D_FILE_OFFSET_BITS=64 -Wall -Winvalid-pch -Wextra -Werror -O2 -g -include rte_config.h -Wcast-qual -Wdeprecated -Wformat -Wformat-nonliteral -Wformat-security -Wmissing-declarations -Wmissing-prototypes -Wnested-externs -Wold-style-definition -Wpointer-arith -Wsign-compare -Wstrict-prototypes -Wundef -Wwrite-strings -Wno-missing-field-initializers -D_GNU_SOURCE -fPIC -march=native -mno-avx512f -mno-avx512f -DALLOW_EXPERIMENTAL_API -DALLOW_INTERNAL_API -fno-strict-aliasing -DRTE_LOG_DEFAULT_LOGTYPE=lib.vhost -DRTE_ANNOTATE_LOCKS -Wthread-safety -MD -MQ lib/librte_vhost.a.p/vhost_iotlb.c.o -MF lib/librte_vhost.a.p/vhost_iotlb.c.o.d -o lib/librte_vhost.a.p/vhost_iotlb.c.o -c ../../root/dpdk/lib/vhost/iotlb.c In file included from ../../root/dpdk/lib/vhost/iotlb.c:11: In file included from ../../root/dpdk/lib/vhost/iotlb.h:10: ../../root/dpdk/lib/vhost/vhost.h:518:2: error: GCC does not allow assert_exclusive_lock attribute in this position on a function definition [-Werror,-Wgcc-compat] __rte_assert_exclusive_lock(&vq->access_lock) ^ ../../root/dpdk/lib/eal/include/rte_lock_annotations.h:29:38: note: expanded from macro '__rte_assert_exclusive_lock' __attribute__((assert_exclusive_lock(__VA_ARGS__))) ^ 1 error generated. ``` OS: CentOS Linux 7 clang: 3.4.2 -- You are receiving this mail because: You are the assignee for the bug.
[PATCH] doc: clarify the existing net/ark guide
Add detail for the existing Arkville configurations FX0 and FX1. Corrected minor errors of omission. Signed-off-by: Shepard Siegel --- doc/guides/nics/ark.rst | 18 ++ 1 file changed, 18 insertions(+) diff --git a/doc/guides/nics/ark.rst b/doc/guides/nics/ark.rst index ba00f14e80..edaa02dc96 100644 --- a/doc/guides/nics/ark.rst +++ b/doc/guides/nics/ark.rst @@ -52,6 +52,10 @@ board. While specific capabilities such as number of physical hardware queue-pairs are negotiated; the driver is designed to remain constant over a broad and extendable feature set. +* FPGA Vendors Supported: AMD/Xilinx and Intel +* Number of RX/TX Queue-Pairs: up to 128 +* PCIe Endpoint Technology: Gen3, Gen4, Gen5 + Intentionally, Arkville by itself DOES NOT provide common NIC capabilities such as offload or receive-side scaling (RSS). These capabilities would be viewed as a gate-level "tax" on @@ -303,6 +307,18 @@ ARK PMD supports the following Arkville RTL PCIe instances including: * ``1d6c:101e`` - AR-ARKA-FX1 [Arkville 64B DPDK Data Mover for Agilex R-Tile] * ``1d6c:101f`` - AR-TK242 [2x100GbE Packet Capture Device] +Arkville RTL Core Configurations +- + +Arkville's RTL core may be configured by the user with different +datapath widths to balance throughput against FPGA logic area. The ARK PMD +has introspection on the RTL core configuration and acts accordingly. +All Arkville configurations present identical RTL user-facing AXI stream +interfaces for both AMD/Xilinx and Intel FPGAs. + +* ARK-FX0 - 256-bit 32B datapath (PCIe Gen3, Gen4) +* ARK-FX1 - 512-bit 64B datapath (PCIe Gen3, Gen4, Gen5) + DPDK and Arkville Firmware Versioning - @@ -334,6 +350,8 @@ Supported Features -- * Dynamic ARK PMD extensions +* Dynamic per-queue MBUF (re)sizing up to 32KB +* SR-IOV, VF-based queue-segregation * Multiple receive and transmit queues * Jumbo frames up to 9K * Hardware Statistics -- 2.25.1
[PATCH 1/2] net/ark: add new device to PCIe allowlist
This patch adds the Arkville FX2 device to the PCIe allowlist.
Signed-off-by: Shepard Siegel
---
drivers/net/ark/ark_ethdev.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/ark/ark_ethdev.c b/drivers/net/ark/ark_ethdev.c
index c654a229f7..b2995427c8 100644
--- a/drivers/net/ark/ark_ethdev.c
+++ b/drivers/net/ark/ark_ethdev.c
@@ -99,6 +99,7 @@ static const struct rte_pci_id pci_id_ark_map[] = {
{RTE_PCI_DEVICE(AR_VENDOR_ID, 0x101c)},
{RTE_PCI_DEVICE(AR_VENDOR_ID, 0x101e)},
{RTE_PCI_DEVICE(AR_VENDOR_ID, 0x101f)},
+ {RTE_PCI_DEVICE(AR_VENDOR_ID, 0x1022)},
{.vendor_id = 0, /* sentinel */ },
};
--
2.25.1
[PATCH 2/2] doc: update ark guide to include new PCIe device
Add descriptions of the new FX2 device to the existing devices. Signed-off-by: Shepard Siegel --- doc/guides/nics/ark.rst | 2 ++ 1 file changed, 2 insertions(+) diff --git a/doc/guides/nics/ark.rst b/doc/guides/nics/ark.rst index edaa02dc96..bbd7419d99 100644 --- a/doc/guides/nics/ark.rst +++ b/doc/guides/nics/ark.rst @@ -306,6 +306,7 @@ ARK PMD supports the following Arkville RTL PCIe instances including: * ``1d6c:101c`` - AR-ARK-SRIOV-VF [Arkville Virtual Function] * ``1d6c:101e`` - AR-ARKA-FX1 [Arkville 64B DPDK Data Mover for Agilex R-Tile] * ``1d6c:101f`` - AR-TK242 [2x100GbE Packet Capture Device] +* ``1d6c:1022`` - AR-ARKA-FX2 [Arkville 128B DPDK Data Mover for Agilex] Arkville RTL Core Configurations - @@ -318,6 +319,7 @@ interfaces for both AMD/Xilinx and Intel FPGAs. * ARK-FX0 - 256-bit 32B datapath (PCIe Gen3, Gen4) * ARK-FX1 - 512-bit 64B datapath (PCIe Gen3, Gen4, Gen5) +* ARK-FX2 - 1024-bit 128B datapath (PCIe Gen5x16 Only) DPDK and Arkville Firmware Versioning - -- 2.25.1
