[dpdk-dev] [PATCH] net/hinic: fix coredump when secondary process using the hinic port.
>From 2eb6ad4d30af54f309f35f16a39a119aa26e7911 Mon Sep 17 00:00:00 2001
From: liqingqing
Date: Sat, 8 Aug 2020 00:03:25 -0400
Subject: [PATCH] fix coredump when secondary process using the hinic port. the
reason is that during the stage of secondary process port initialization, it
lack the initialization of "eth_dev->dev_ops".
Signed-off-by: liqingqing
---
drivers/net/hinic/hinic_pmd_ethdev.c | 21 -
1 file changed, 12 insertions(+), 9 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c
b/drivers/net/hinic/hinic_pmd_ethdev.c
index 67e6afcf7..26c6098d3 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -3060,15 +3060,6 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
int rc;
pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
-
- /* EAL is SECONDARY and eth_dev is already created */
- if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
- PMD_DRV_LOG(INFO, "Initialize %s in secondary process",
- eth_dev->data->name);
-
- return 0;
- }
-
nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
memset(nic_dev, 0, sizeof(*nic_dev));
@@ -3206,6 +3197,18 @@ static int hinic_dev_init(struct rte_eth_dev *eth_dev)
eth_dev->rx_pkt_burst = hinic_recv_pkts;
eth_dev->tx_pkt_burst = hinic_xmit_pkts;
+ /* EAL is SECONDARY and eth_dev is already created */
+ if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
+ PMD_DRV_LOG(INFO, "Initialize %s in secondary process",
eth_dev->data->name);
+
+ struct hinic_nic_dev *nic_dev =
HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
+ if (HINIC_IS_VF(nic_dev->hwdev))
+ eth_dev->dev_ops = &hinic_pmd_vf_ops;
+ else
+ eth_dev->dev_ops = &hinic_pmd_ops;
+ return 0;
+ }
+
return hinic_func_init(eth_dev);
}
--
2.19.1
[dpdk-dev] [PATCH v2] net/hinic: fix secondary process's using coredump.
fix coredump when secondary process using the hinic port.
the reason is that during the stage of
secondary process port initialization,
it lack the initialization of "eth_dev->dev_ops".
Signed-off-by: liqingqing
---
v2: solve the coding style issue.
---
drivers/net/hinic/hinic_pmd_ethdev.c | 21 -
1 file changed, 12 insertions(+), 9 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c
b/drivers/net/hinic/hinic_pmd_ethdev.c
index 67e6afcf7..26c6098d3 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -3060,15 +3060,6 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
int rc;
pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
-
- /* EAL is SECONDARY and eth_dev is already created */
- if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
- PMD_DRV_LOG(INFO, "Initialize %s in secondary process",
- eth_dev->data->name);
-
- return 0;
- }
-
nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
memset(nic_dev, 0, sizeof(*nic_dev));
@@ -3206,6 +3197,18 @@ static int hinic_dev_init(struct rte_eth_dev *eth_dev)
eth_dev->rx_pkt_burst = hinic_recv_pkts;
eth_dev->tx_pkt_burst = hinic_xmit_pkts;
+ /* EAL is SECONDARY and eth_dev is already created */
+ if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
+ PMD_DRV_LOG(INFO, "Initialize %s in secondary process",
eth_dev->data->name);
+
+ struct hinic_nic_dev *nic_dev =
HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
+ if (HINIC_IS_VF(nic_dev->hwdev))
+ eth_dev->dev_ops = &hinic_pmd_vf_ops;
+ else
+ eth_dev->dev_ops = &hinic_pmd_ops;
+ return 0;
+ }
+
return hinic_func_init(eth_dev);
}
--
2.19.1
[dpdk-dev] [PATCH v2] net/tap: free mempool when closing
From: Yunjian Wang
When setup tx queues, we will create a mempool for the 'gso_ctx'.
The mempool is not freed when closing tap device. If free the tap
device and create it with different name, it will create a new
mempool. This maybe cause an OOM.
The snprintf function return value is not checked and the mempool
name may be truncated. This patch also fix it.
Fixes: 050316a88313 ("net/tap: support TSO (TCP Segment Offload)")
Cc: sta...@dpdk.org
Signed-off-by: Yunjian Wang
---
v2: Updated the error log message as suggested by Ferruh Yigit
---
drivers/net/tap/rte_eth_tap.c | 43 +--
drivers/net/tap/rte_eth_tap.h | 1 +
2 files changed, 27 insertions(+), 17 deletions(-)
diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
index 339f24bf8..185de8aee 100644
--- a/drivers/net/tap/rte_eth_tap.c
+++ b/drivers/net/tap/rte_eth_tap.c
@@ -1070,6 +1070,9 @@ tap_dev_close(struct rte_eth_dev *dev)
&internals->remote_initial_flags);
}
+ rte_mempool_free(internals->gso_ctx_mp);
+ internals->gso_ctx_mp = NULL;
+
if (internals->ka_fd != -1) {
close(internals->ka_fd);
internals->ka_fd = -1;
@@ -1317,26 +1320,31 @@ tap_gso_ctx_setup(struct rte_gso_ctx *gso_ctx, struct
rte_eth_dev *dev)
{
uint32_t gso_types;
char pool_name[64];
-
- /*
-* Create private mbuf pool with TAP_GSO_MBUF_SEG_SIZE bytes
-* size per mbuf use this pool for both direct and indirect mbufs
-*/
-
- struct rte_mempool *mp; /* Mempool for GSO packets */
+ struct pmd_internals *pmd = dev->data->dev_private;
+ int ret;
/* initialize GSO context */
gso_types = DEV_TX_OFFLOAD_TCP_TSO;
- snprintf(pool_name, sizeof(pool_name), "mp_%s", dev->device->name);
- mp = rte_mempool_lookup((const char *)pool_name);
- if (!mp) {
- mp = rte_pktmbuf_pool_create(pool_name, TAP_GSO_MBUFS_NUM,
- TAP_GSO_MBUF_CACHE_SIZE, 0,
+ if (!pmd->gso_ctx_mp) {
+ /*
+* Create private mbuf pool with TAP_GSO_MBUF_SEG_SIZE
+* bytes size per mbuf use this pool for both direct and
+* indirect mbufs
+*/
+ ret = snprintf(pool_name, sizeof(pool_name), "mp_%s",
+ dev->device->name);
+ if (ret < 0 || ret >= (int)sizeof(pool_name)) {
+ TAP_LOG(ERR,
+ "%s: failed to create mbuf pool name for device
%s,"
+ "device name too long or output error, ret:
%d\n",
+ pmd->name, dev->device->name, ret);
+ return -ENAMETOOLONG;
+ }
+ pmd->gso_ctx_mp = rte_pktmbuf_pool_create(pool_name,
+ TAP_GSO_MBUFS_NUM, TAP_GSO_MBUF_CACHE_SIZE, 0,
RTE_PKTMBUF_HEADROOM + TAP_GSO_MBUF_SEG_SIZE,
SOCKET_ID_ANY);
- if (!mp) {
- struct pmd_internals *pmd = dev->data->dev_private;
-
+ if (!pmd->gso_ctx_mp) {
TAP_LOG(ERR,
"%s: failed to create mbuf pool for device
%s\n",
pmd->name, dev->device->name);
@@ -1344,8 +1352,8 @@ tap_gso_ctx_setup(struct rte_gso_ctx *gso_ctx, struct
rte_eth_dev *dev)
}
}
- gso_ctx->direct_pool = mp;
- gso_ctx->indirect_pool = mp;
+ gso_ctx->direct_pool = pmd->gso_ctx_mp;
+ gso_ctx->indirect_pool = pmd->gso_ctx_mp;
gso_ctx->gso_types = gso_types;
gso_ctx->gso_size = 0; /* gso_size is set in tx_burst() per packet */
gso_ctx->flag = 0;
@@ -1842,6 +1850,7 @@ eth_dev_tap_create(struct rte_vdev_device *vdev, const
char *tap_name,
pmd->type = type;
pmd->ka_fd = -1;
pmd->nlsk_fd = -1;
+ pmd->gso_ctx_mp = NULL;
pmd->ioctl_sock = socket(AF_INET, SOCK_DGRAM, 0);
if (pmd->ioctl_sock == -1) {
diff --git a/drivers/net/tap/rte_eth_tap.h b/drivers/net/tap/rte_eth_tap.h
index 8d6d53dc0..ba45de840 100644
--- a/drivers/net/tap/rte_eth_tap.h
+++ b/drivers/net/tap/rte_eth_tap.h
@@ -91,6 +91,7 @@ struct pmd_internals {
struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES]; /* List of TX queues */
struct rte_intr_handle intr_handle; /* LSC interrupt handle. */
int ka_fd;/* keep-alive file descriptor */
+ struct rte_mempool *gso_ctx_mp; /* Mempool for GSO packets */
};
struct pmd_process_private {
--
2.23.0
Re: [dpdk-dev] [EXT] [dpdk-announce] release candidate 20.08-rc4
Hi, Tested on NXP platforms (DPAA & DPAA2) and throughput numbers are as expected. OS: Ubuntu 18.04 Platforms: LS1046ARDB LS1043ARDB LS2088ARDB LX2160ARDB Tested applications: l2fwd l3fwd ipsec-secgw Thanks & Regards Apeksha > -Original Message- > From: dev On Behalf Of Thomas Monjalon > Sent: Thursday, August 6, 2020 6:32 AM > To: annou...@dpdk.org > Subject: [EXT] [dpdk-dev] [dpdk-announce] release candidate 20.08-rc4 > > Caution: EXT Email > > A new DPDK release candidate is ready for testing: > > https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgit.dp > dk.org%2Fdpdk%2Ftag%2F%3Fid%3Dv20.08- > rc4&data=02%7C01%7Capeksha.gupta%40nxp.com%7C7c15cea37f7745 > 696ddd08d839a4623f%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0% > 7C637322725471324274&sdata=DMaY8yWZqiy1i3konzwuKuAoVjW1wy > KRfxT3w%2Fkuh6Q%3D&reserved=0 > > There are 27 new patches in this snapshot. > > Release notes: > > https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fdoc.dp > dk.org%2Fguides%2Frel_notes%2Frelease_20_08.html&data=02%7C01 > %7Capeksha.gupta%40nxp.com%7C7c15cea37f7745696ddd08d839a4623f%7 > C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C637322725471324274 > &sdata=YEzSC2QHzODZgO5ClnWmYckNSeUZ663QwIJQtUTpYhQ%3D&a > mp;reserved=0 > More deprecation notices are pending for acceptance. > > This is the last release candidate for DPDK 20.08. > After -rc4 there should be only doc updates. > Please review pending patches related to the documentation > and the deprecation notices: > > https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpatch > es.dpdk.org%2Fproject%2Fdpdk%2Flist%2F%3Fq%3Ddoc&data=02%7C0 > 1%7Capeksha.gupta%40nxp.com%7C7c15cea37f7745696ddd08d839a4623f% > 7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C637322725471324274 > &sdata=iBnru8Y2PVjqMZVAm68pfUvDtOw08deX1FiX9Lqfp4g%3D& > reserved=0 > > Please share some release validation results > by replying to this message (at dev@dpdk.org). > If no critical issue is reported, 20.08 will be closed end of this week. > > If you are preparing the next release cycle, please send your v1 patches > before the end of August. > We should also build and discuss the features roadmap for 20.11 soon, > so we can prioritize tasks and avoid flooding maintainers. > > Thank you everyone >
[dpdk-dev] [PATCH v2 1/2] ethdev: add level support for RSS offload types
From: Kiran Kumar K
This patch reserves 2 bits as input selection to select Inner and
outer layers for RSS computation. It is combined with existing
ETH_RSS_* to choose Inner or outer layers for L2, L3 and L4.
This functionality already exists in rte_flow through level parameter in
RSS action configuration rte_flow_action_rss.
Signed-off-by: Kiran Kumar K
---
v2 changes:
* Reserved bit 50 & 51
lib/librte_ethdev/rte_ethdev.h | 27 +++
1 file changed, 27 insertions(+)
diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
index d29930fd8..28184cc85 100644
--- a/lib/librte_ethdev/rte_ethdev.h
+++ b/lib/librte_ethdev/rte_ethdev.h
@@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
#define RTE_ETH_RSS_L3_PRE64 (1ULL << 53)
#define RTE_ETH_RSS_L3_PRE96 (1ULL << 52)
+/*
+ * We use the following macros to combine with the above layers to choose
+ * inner and outer layers or both for RSS computation.
+ * Note: Default is 0: inner layers, 1: outer layers, 2: both
+ * bit 50 and 51 are reserved for this.
+ */
+
+/**
+ * Level 0, It basically stands for the innermost encapsulation level RSS
+ * can be performed on according to PMD and device capabilities.
+ */
+#define ETH_RSS_LEVEL_INNER(0ULL << 50)
+/**
+ * Level 1, It basically stands for the outermost encapsulation level RSS
+ * can be performed on according to PMD and device capabilities.
+ */
+#define ETH_RSS_LEVEL_OUTER(1ULL << 50)
+/**
+ * Level 2, It basically stands for the both inner and outermost
+ * encapsulation level RSS can be performed on according to PMD and
+ * device capabilities.
+ */
+#define ETH_RSS_LEVEL_INNER_OUTER (2ULL << 50)
+#define ETH_RSS_LEVEL_MASK(3ULL << 50)
+
+#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
+
/**
* For input set change of hash filter, if SRC_ONLY and DST_ONLY of
* the same level are used simultaneously, it is the same case as
--
2.25.1
[dpdk-dev] [PATCH v2 2/2] net/octeontx2: add rss hash level support
From: Kiran Kumar K
Add support to choose rss hash level from ethdev rss config.
Signed-off-by: Kiran Kumar K
---
drivers/net/octeontx2/otx2_ethdev.h | 4 +++-
drivers/net/octeontx2/otx2_rss.c| 9 +++--
2 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/drivers/net/octeontx2/otx2_ethdev.h
b/drivers/net/octeontx2/otx2_ethdev.h
index e9efe52bb..953445ecb 100644
--- a/drivers/net/octeontx2/otx2_ethdev.h
+++ b/drivers/net/octeontx2/otx2_ethdev.h
@@ -119,7 +119,9 @@
#define NIX_RSS_OFFLOAD(ETH_RSS_PORT | ETH_RSS_IP |
ETH_RSS_UDP |\
ETH_RSS_TCP | ETH_RSS_SCTP | \
ETH_RSS_TUNNEL | ETH_RSS_L2_PAYLOAD | \
-NIX_RSS_L3_L4_SRC_DST)
+NIX_RSS_L3_L4_SRC_DST | ETH_RSS_LEVEL_INNER | \
+ETH_RSS_LEVEL_OUTER | \
+ETH_RSS_LEVEL_INNER_OUTER)
#define NIX_TX_OFFLOAD_CAPA ( \
DEV_TX_OFFLOAD_MBUF_FAST_FREE | \
diff --git a/drivers/net/octeontx2/otx2_rss.c b/drivers/net/octeontx2/otx2_rss.c
index d859937e6..194f4e8af 100644
--- a/drivers/net/octeontx2/otx2_rss.c
+++ b/drivers/net/octeontx2/otx2_rss.c
@@ -323,6 +323,7 @@ otx2_nix_rss_hash_update(struct rte_eth_dev *eth_dev,
struct rte_eth_rss_conf *rss_conf)
{
struct otx2_eth_dev *dev = otx2_eth_pmd_priv(eth_dev);
+ uint8_t rss_hash_level;
uint32_t flowkey_cfg;
uint8_t alg_idx;
int rc;
@@ -339,7 +340,9 @@ otx2_nix_rss_hash_update(struct rte_eth_dev *eth_dev,
otx2_nix_rss_set_key(dev, rss_conf->rss_key,
(uint32_t)rss_conf->rss_key_len);
- flowkey_cfg = otx2_rss_ethdev_to_nix(dev, rss_conf->rss_hf, 0);
+ rss_hash_level = ETH_RSS_LEVEL(rss_conf->rss_hf);
+ flowkey_cfg =
+ otx2_rss_ethdev_to_nix(dev, rss_conf->rss_hf, rss_hash_level);
rc = otx2_rss_set_hf(dev, flowkey_cfg, &alg_idx,
NIX_DEFAULT_RSS_CTX_GROUP,
@@ -375,6 +378,7 @@ otx2_nix_rss_config(struct rte_eth_dev *eth_dev)
{
struct otx2_eth_dev *dev = otx2_eth_pmd_priv(eth_dev);
uint32_t idx, qcnt = eth_dev->data->nb_rx_queues;
+ uint8_t rss_hash_level;
uint32_t flowkey_cfg;
uint64_t rss_hf;
uint8_t alg_idx;
@@ -399,7 +403,8 @@ otx2_nix_rss_config(struct rte_eth_dev *eth_dev)
}
rss_hf = eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf;
- flowkey_cfg = otx2_rss_ethdev_to_nix(dev, rss_hf, 0);
+ rss_hash_level = ETH_RSS_LEVEL(rss_hf);
+ flowkey_cfg = otx2_rss_ethdev_to_nix(dev, rss_hf, rss_hash_level);
rc = otx2_rss_set_hf(dev, flowkey_cfg, &alg_idx,
NIX_DEFAULT_RSS_CTX_GROUP,
--
2.25.1
[dpdk-dev] [PATCH] lib/metrics: fix memory leak
fix memory leak
Fixes: c5b7197f66 ("telemetry: move some functions to metrics library")
Signed-off-by: Gaurav Singh
---
lib/librte_metrics/rte_metrics_telemetry.c | 21 -
1 file changed, 16 insertions(+), 5 deletions(-)
diff --git a/lib/librte_metrics/rte_metrics_telemetry.c
b/lib/librte_metrics/rte_metrics_telemetry.c
index 289ebae0b..7b6d1063c 100644
--- a/lib/librte_metrics/rte_metrics_telemetry.c
+++ b/lib/librte_metrics/rte_metrics_telemetry.c
@@ -41,12 +41,17 @@ rte_metrics_tel_reg_port_ethdev_to_metrics(uint16_t port_id)
}
xstats_names = malloc(sizeof(*xstats_names) * num_xstats);
+ if (xstats_names == NULL) {
+ METRICS_LOG_ERR("Failed to malloc memory for xstats_names");
+ return -ENOMEM;
+ }
+
eth_xstats_names = malloc(sizeof(struct rte_eth_xstat_name)
* num_xstats);
- if (eth_xstats_names == NULL || xstats_names == NULL) {
- METRICS_LOG_ERR("Failed to malloc memory for xstats_names");
- ret = -ENOMEM;
- goto free_xstats;
+ if (eth_xstats_names == NULL) {
+ METRICS_LOG_ERR("Failed to malloc memory for eth_xstats_names");
+ free(xstats_names);
+ return -ENOMEM;
}
if (rte_eth_xstats_get_names(port_id,
@@ -167,9 +172,15 @@ rte_metrics_tel_format_port(uint32_t pid, json_t *ports,
}
metrics = malloc(sizeof(struct rte_metric_value) * num_metrics);
+ if (metrics == NULL) {
+ METRICS_LOG_ERR("Cannot allocate memory");
+ return -ENOMEM;
+ }
+
names = malloc(sizeof(struct rte_metric_name) * num_metrics);
- if (metrics == NULL || names == NULL) {
+ if (names == NULL) {
METRICS_LOG_ERR("Cannot allocate memory");
+ free(metrics);
return -ENOMEM;
}
--
2.17.1
Re: [dpdk-dev] [PATCH v2 1/2] ethdev: add level support for RSS offload types
On Sat, Aug 8, 2020 at 7:36 AM wrote:
> From: Kiran Kumar K
>
> This patch reserves 2 bits as input selection to select Inner and
> outer layers for RSS computation. It is combined with existing
> ETH_RSS_* to choose Inner or outer layers for L2, L3 and L4.
How do you plan to use this? Do you need to make any changes to testpmd?
> This functionality already exists in rte_flow through level parameter in
> RSS action configuration rte_flow_action_rss.
>
> Signed-off-by: Kiran Kumar K
> ---
> v2 changes:
> * Reserved bit 50 & 51
>
> lib/librte_ethdev/rte_ethdev.h | 27 +++
> 1 file changed, 27 insertions(+)
>
> diff --git a/lib/librte_ethdev/rte_ethdev.h
> b/lib/librte_ethdev/rte_ethdev.h
> index d29930fd8..28184cc85 100644
> --- a/lib/librte_ethdev/rte_ethdev.h
> +++ b/lib/librte_ethdev/rte_ethdev.h
> @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
> #define RTE_ETH_RSS_L3_PRE64 (1ULL << 53)
> #define RTE_ETH_RSS_L3_PRE96 (1ULL << 52)
>
> +/*
> + * We use the following macros to combine with the above layers to choose
> + * inner and outer layers or both for RSS computation.
> + * Note: Default is 0: inner layers, 1: outer layers, 2: both
> + * bit 50 and 51 are reserved for this.
> + */
> +
> +/**
> + * Level 0, It basically stands for the innermost encapsulation level RSS
> + * can be performed on according to PMD and device capabilities.
> + */
> +#define ETH_RSS_LEVEL_INNER(0ULL << 50)
> +/**
> + * Level 1, It basically stands for the outermost encapsulation level RSS
> + * can be performed on according to PMD and device capabilities.
> + */
> +#define ETH_RSS_LEVEL_OUTER(1ULL << 50)
> +/**
> + * Level 2, It basically stands for the both inner and outermost
> + * encapsulation level RSS can be performed on according to PMD and
> + * device capabilities.
> + */
> +#define ETH_RSS_LEVEL_INNER_OUTER (2ULL << 50)
> +#define ETH_RSS_LEVEL_MASK(3ULL << 50)
> +
> +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
> +
> /**
> * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
> * the same level are used simultaneously, it is the same case as
> --
> 2.25.1
>
>
[dpdk-dev] [dpdk-announce] DPDK 20.08 released
A new release is available: https://fast.dpdk.org/rel/dpdk-20.08.tar.xz This version could be named Twins: 20.08 is released on 2020-08-08. The other numbers are: 1175 commits from 159 authors 1199 files changed, 137604 insertions(+), 35611 deletions(-) It is not planned to start a maintenance branch for 20.08. This version as previous ones are ABI-compatible with 19.11. Below are some new features, grouped by category. * General - external thread registration API - bit operations API for drivers - VFIO PF with VF token - new mempool ring modes (RTS/HTS) * Networking - eCPRI offload with flow API - Tx scheduling offload * Cryptography - crypto-CRC chained operation for DOCSIS protocol * RegEx - new device class API for future RegEx drivers - Nvidia Mellanox RegEx driver for BlueField-2 DPU * Tools - testpmd 5-tuple swap for L2/L3/L4 - performance test application for flow rules - l2fwd forwarding between asymmetric ports More details in the release notes: http://doc.dpdk.org/guides/rel_notes/release_20_08.html There are 52 new contributors (including authors, reviewers and testers). Welcome to Adrián Moreno, Archit Pandey, Bin Huang, Devendra Singh Rawat, Dodji Seketeli, Dongyang Pan, Evgeny Efimov, Francis Kelly, Gaurav Singh, Gregory Etelson, Guy Kaneti, Hongzhi Guo, Hrvoje Habjanic, Jakub Chylkowski, Jecky Pei, Jeff Kirsher, Kamil Bednarczyk, Levend Sayar, Long Li, Lotem Leder, Maciej Hefczyc, Mao Jiang, Mateusz Kowalski, Netanel Gonen, Nick Nunley, Nir Efrati, Parav Pandit, Patrick Fu, Pavel Ivashchenko, Piotr Skajewski, Renata Saiakhova, Roman Fridlyand, Roman Kapl, Sasha Neftin, Sergey Madaminov, Shibin Koikkara Reeny, Shuanglin Wang, Shy Shyman, Simon Horman, Stanislaw Grzeszczak, Surabhi Boob, Thierry Martin, Timothy McDaniel, Todd Fujinaka, Vitaly Lifshits, Vivien Didelot, Wei Xie, Weifeng Li, Yicai Lu, Yuval Avnery, Yuying Zhang, and Zhiguang He. Below is the number of commits per employer (with authors count): 362 Intel (45) 276 Nvidia Mellanox (24) 157 Broadcom (15) 68 Huawei (10) 64 Red Hat (5) 51 NXP (8) 50 ARM (5) 49 Marvell (12) 32 Microsoft (3) 16 BIFIT (1) 7 Emumba (2) 6 Solarflare (1) 6 Chelsio (2) 5 OKTET Labs (4) The new features for 20.11 can be submitted during one month. DPDK 20.11 will be a big and busy release. Please check the schedule: http://core.dpdk.org/roadmap#dates Thanks everyone, stay safe and keep up morale.
[dpdk-dev] [PATCH] Virtio TX: reverting a small change causing Virtio standard TX broken
Virtio Standard TX broken : Reverting a small change added few months back
which has caused breakage in Virtio Standard TX path.
During basic ping Tx/Rx testing of net_virtio pmd driver using standard Tx
path, We saw TX is broken since anomaly code added by below patch once submitted
http://git.dpdk.org/dpdk/patch/?id=57f90f89458807bccc63425e4b72796870177977
---
drivers/net/virtio/virtqueue.h | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h
index 105a9c00c..81118d9fb 100644
--- a/drivers/net/virtio/virtqueue.h
+++ b/drivers/net/virtio/virtqueue.h
@@ -607,10 +607,8 @@ virtqueue_notify(struct virtqueue *vq)
/* avoid write operation when necessary, to lessen cache issues */
#define ASSIGN_UNLESS_EQUAL(var, val) do { \
- typeof(var) var_ = (var); \
- typeof(val) val_ = (val); \
- if ((var_) != (val_)) \
- (var_) = (val_);\
+ if ((var) != (val)) \
+ (var) = (val); \
} while (0)
#define virtqueue_clear_net_hdr(hdr) do { \
--
2.28.0.windows.1
