Re: [RFC] ethdev: introduce maximum Rx buffer size

[lihuisong (C)]

Hi Andrew,
Thanks for your review.


在 2023/8/11 20:07, Andrew Rybchenko 写道:

On 8/8/23 07:02, Huisong Li wrote:

The Rx buffer size stands for the size hardware supported to receive
packets in one mbuf. The "min_rx_bufsize" is the minimum buffer hardware
supported in Rx. Actually, some engines also have the maximum buffer
specification, like, hns3. For these engines, the available data size
of one mbuf in Rx also depends on the maximum buffer hardware supported.
So introduce maximum Rx buffer size in struct rte_eth_dev_info to report
user to avoid memory waste.


I think that the field should be defined as for informational purposes
only (highlighted in comments). I.e. if application specifies larger Rx
buffer, driver should accept it and just pass smaller value value to HW.

Ok, will add it.

Also I think it would be useful to log warning from Rx queue setup
if provided Rx buffer is larger than maximum reported by the driver.

Ack




Signed-off-by: Huisong Li 
---
  lib/ethdev/rte_ethdev.c | 1 +
  lib/ethdev/rte_ethdev.h | 4 ++--
  2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index 0840d2b594..6d1b92e607 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -3689,6 +3689,7 @@ rte_eth_dev_info_get(uint16_t port_id, struct 
rte_eth_dev_info *dev_info)

  dev_info->min_mtu = RTE_ETHER_MIN_LEN - RTE_ETHER_HDR_LEN -
  RTE_ETHER_CRC_LEN;
  dev_info->max_mtu = UINT16_MAX;
+    dev_info->max_rx_bufsize = UINT32_MAX;
    if (*dev->dev_ops->dev_infos_get == NULL)
  return -ENOTSUP;
diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
index 04a2564f22..1f0ab9c5d8 100644
--- a/lib/ethdev/rte_ethdev.h
+++ b/lib/ethdev/rte_ethdev.h
@@ -1779,8 +1779,8 @@ struct rte_eth_dev_info {
  struct rte_eth_switch_info switch_info;
  /** Supported error handling mode. */
  enum rte_eth_err_handle_mode err_handle_mode;
-
-    uint64_t reserved_64s[2]; /**< Reserved for future fields */
+    uint32_t max_rx_bufsize; /**< Maximum size of Rx buffer. */


IMHO, comment should be aligned similar to comments below.
Since the next release is ABI breaking, I think it should be put
nearby min_rx_bufsize to make it easier to notice it.

Yes, let's put min/max_rx_bufsize together.



+    uint32_t reserved_32s[3]; /**< Reserved for future fields */
  void *reserved_ptrs[2];   /**< Reserved for future fields */
  };


.

Intel Downfall mitigation

[Thomas Monjalon]

Hello all,

Regarding the vulnerability recently discovered on Intel AVX CPU:
https://downfall.page/
Some microcode fixes are published by Intel.

Should we mitigate the vulnerability in DPDK as well?
In general, AVX512 is disabled by default in DPDK.
Other AVX sizes are enabled if possible.

The memcpy 512 is enabled at compilation only with RTE_MEMCPY_AVX512.
For now, there is no switch for other memcpy paths.

There is a method to disable most SIMD code at runtime.
>From the application before EAL init:
rte_vect_set_max_simd_bitwidth(RTE_VECT_SIMD_DISABLED)
or from user input:
--force-max-simd-bitwidth=64
As far as I know it will disable any SIMD code (including AVX)
in most code paths except for rte_memcpy().

Opinions, comments?

Re: 21.11.5 patches review and test

[YangHang Liu]

Hi, Kevin

RedHat QE did not find any new issues about 21.11.5 rc1 dpdk during the
tests.

I tested below 18 scenarios and all got PASS on RHEL9:

   - Guest with device assignment(PF) throughput testing(1G hugepage size):
   PASS
   - Guest with device assignment(PF) throughput testing(2M hugepage size)
   : PASS
   - Guest with device assignment(VF) throughput testing: PASS
   - PVP (host dpdk testpmd as vswitch) 1Q: throughput testing: PASS
   - PVP vhost-user 2Q throughput testing: PASS
   - PVP vhost-user 1Q - cross numa node throughput testing: PASS
   - Guest with vhost-user 2 queues throughput testing: PASS
   - vhost-user reconnect with dpdk-client, qemu-server qemu reconnect: PASS
   - vhost-user reconnect with dpdk-client, qemu-server ovs reconnect: PASS
   - PVP  reconnect with dpdk-client, qemu-server: PASS
   - PVP 1Q live migration testing: PASS
   - PVP 1Q cross numa node live migration testing: PASS
   - Guest with ovs+dpdk+vhost-user 1Q live migration testing: PASS
   - Guest with ovs+dpdk+vhost-user 1Q live migration testing (2M): PASS
   - Guest with ovs+dpdk+vhost-user 2Q live migration testing: PASS
   - Guest with ovs+dpdk+vhost-user 4Q live migration testing: PASS
   - Host PF + DPDK testing: PASS
   - Host VF + DPDK testing: PASS

Test Versions:

   - qemu-kvm-6.2.0
   - kernel 5.14
   - dpdk 21.11.5-rc1

  commit 89133ca5b161036ace8fee04dd39ca94533fffc9

  Author: Kevin Traynor 

  Date:   Wed Aug 2 13:23:13 2023 +0100

  version: 21.11.5-rc1
  Signed-off-by: Kevin Traynor <
ktray...@redhat.com>

   - Test device : X540-AT2 NIC(ixgbe, 10G)


Best Regards,
YangHang Liu




On Thu, Aug 3, 2023 at 12:31 AM Kevin Traynor  wrote:

> Hi all,
>
> Here is a list of patches targeted for stable release 21.11.5.
>
> The planned date for the final release is 22nd August.
>
> Please help with testing and validation of your use cases and report
> any issues/results with reply-all to this mail. For the final release
> the fixes and reported validations will be added to the release notes.
>
> A release candidate tarball can be found at:
>
> https://dpdk.org/browse/dpdk-stable/tag/?id=v21.11.5-rc1
>
> These patches are located at branch 21.11 of dpdk-stable repo:
> https://dpdk.org/browse/dpdk-stable/
>
> Thanks.
>
> Kevin
>
> ---
> Aakash Sasidharan (1):
>   test/crypto: fix PDCP-SDAP test vectors
>
> Akhil Goyal (1):
>   doc: fix auth algos in cryptoperf app
>
> Alexander Kozyrev (1):
>   net/mlx5: forbid MPRQ restart
>
> Ali Alnubani (1):
>   doc: fix typos and wording in flow API guide
>
> Artemii Morozov (1):
>   common/sfc_efx/base: fix Rx queue without RSS hash prefix
>
> Ashwin Sekhar T K (1):
>   doc: fix typo in graph guide
>
> Bing Zhao (2):
>   net/mlx5: fix drop action memory leak
>   net/mlx5: fix validation for conntrack indirect action
>
> Boleslav Stankevich (2):
>   net/virtio: propagate interrupt configuration error values
>   net/virtio: fix initialization to return negative errno
>
> Bruce Richardson (5):
>   kernel/freebsd: fix function parameter list
>   build: fix case of project language name
>   telemetry: fix autotest on Alpine
>   eal: avoid calling cleanup twice
>   test/bonding: fix include of standard header
>
> Chaoyong He (1):
>   net/nfp: fix offloading flows
>
> Chengwen Feng (4):
>   net/hns3: fix Rx multiple firmware reset interrupts
>   net/hns3: fix mbuf leakage when RxQ started during reset
>   net/hns3: fix mbuf leakage when RxQ started after reset
>   net/hns3: fix device start return value
>
> Ciara Power (3):
>   crypto/scheduler: fix last element for valid args
>   crypto/ipsec_mb: optimize allocation in session
>   app/crypto-perf: fix socket ID default value
>
> Dariusz Sosnowski (1):
>   net/mlx5: fix drop action attribute validation
>
> David Christensen (1):
>   net/tap: set locally administered bit for fixed MAC address
>
> David Marchand (6):
>   net/mlx5: enhance error log for tunnel offloading
>   net/virtio-user: fix leak when initialisation fails
>   examples/l2fwd-cat: fix external build
>   test: add graph tests
>   mbuf: fix Doxygen comment of distributor metadata
>   ci: fix libabigail cache in GHA
>
> Dengdui Huang (3):
>   net/hns3: fix variable type mismatch
>   net/hns3: fix inaccurate log
>   net/hns3: fix redundant line break in log
>
> Denis Pryazhennikov (3):
>   ethdev: update documentation for API to set FEC
>   ethdev: check that at least one FEC mode is specified
>   ethdev: update documentation for API to get FEC
>
> Devendra Singh Rawat (1):
>   net/qede: fix RSS indirection table initialization
>
> Didier Pallard (1):
>   crypto/openssl: skip workaround at compilation time
>
> Dongdong Liu (3):
>   doc: fix kernel patch link in hns3 

[PATCH] doc: add NVIDIA BlueField-3 reference link

[Thomas Monjalon]

A link to the detailed guide of BlueField-3 hardware on NVIDIA website
is added to the BlueField platform guide.

Signed-off-by: Thomas Monjalon 
---
 doc/guides/platform/bluefield.rst | 1 +
 1 file changed, 1 insertion(+)

diff --git a/doc/guides/platform/bluefield.rst 
b/doc/guides/platform/bluefield.rst
index 98df515241..322b08a217 100644
--- a/doc/guides/platform/bluefield.rst
+++ b/doc/guides/platform/bluefield.rst
@@ -19,6 +19,7 @@ Supported BlueField Platforms
 -
 
 - `BlueField-2 
`_
+- `BlueField-3 
`_
 
 
 Common Offload HW Drivers
-- 
2.41.0

Re: [PATCH] devtools: list symbols by version

[David Marchand]

On Mon, Aug 14, 2023 at 11:49 AM Thomas Monjalon  wrote:
>
> vendredi 11 août 2023, David Marchand:
> > Make it easier to list experimental symbols added in a certain version.
> > While at it, add a check on map symbol files content to avoid breaking
> > this listing tool.
>
> Is there a relation between the new check and the new -V option?
> I feel it would be clearer in 2 separate patches.

The new option relies on a strict format in the .map files.
I could add the check first in a separate patch if you prefer.


-- 
David Marchand

[PATCH v2 1/2] devtools: add check on symbol maps format

[David Marchand]

Add a check on symbol maps format (fixing ethdev at the same time).
This will be required by a next commit.

Signed-off-by: David Marchand 
---
Changes since v1:
- moved this check in a separate patch,
- fixed ethdev map file,

---
 devtools/check-symbol-maps.sh | 21 +
 lib/ethdev/version.map| 30 ++
 2 files changed, 35 insertions(+), 16 deletions(-)

diff --git a/devtools/check-symbol-maps.sh b/devtools/check-symbol-maps.sh
index 8c116bfa9c..1ad48d9243 100755
--- a/devtools/check-symbol-maps.sh
+++ b/devtools/check-symbol-maps.sh
@@ -74,4 +74,25 @@ if [ -n "$empty_maps" ] ; then
 ret=1
 fi
 
+find_bad_format_maps ()
+{
+for map in $@ ; do
+cat $map | awk '
+/^(DPDK_[0-9]*|EXPERIMENTAL|INTERNAL) {$/ { next; } # start of a 
section
+/^};$/ { next; } # end of a section
+/^$/ { next; } # empty line
+/^\t(global:|local: \*;)$/ { next; } # qualifiers
+/^\t[a-zA-Z_0-9]*;(| # WINDOWS_NO_EXPORT)$/ { next; } # symbols
+/^\t# added in [0-9]*\.[0-9]*$/ { next; } # version comments
+{ print $0; }' || echo $map
+done
+}
+
+bad_format_maps=$(find_bad_format_maps $@)
+if [ -n "$bad_format_maps" ] ; then
+echo "Found badly formatted maps:"
+echo "$bad_format_maps"
+ret=1
+fi
+
 exit $ret
diff --git a/lib/ethdev/version.map b/lib/ethdev/version.map
index b965d6aa52..da4f847116 100644
--- a/lib/ethdev/version.map
+++ b/lib/ethdev/version.map
@@ -181,7 +181,20 @@ EXPERIMENTAL {
__rte_ethdev_trace_tx_burst;
rte_flow_get_aged_flows;
 
-   # Marked as experimental in 20.11
+   # added in 20.11
+   rte_eth_hairpin_bind;
+   rte_eth_hairpin_get_peer_ports;
+   rte_eth_hairpin_unbind;
+   rte_eth_link_speed_to_str;
+   rte_eth_link_to_str;
+   rte_eth_fec_get_capability;
+   rte_eth_fec_get;
+   rte_eth_fec_set;
+   rte_flow_tunnel_decap_set;
+   rte_flow_tunnel_match;
+   rte_flow_get_restore_info;
+   rte_flow_tunnel_action_decap_release;
+   rte_flow_tunnel_item_release;
rte_tm_capabilities_get;
rte_tm_get_number_of_leaf_nodes;
rte_tm_hierarchy_commit;
@@ -213,21 +226,6 @@ EXPERIMENTAL {
rte_tm_wred_profile_add;
rte_tm_wred_profile_delete;
 
-   # added in 20.11
-   rte_eth_hairpin_bind;
-   rte_eth_hairpin_get_peer_ports;
-   rte_eth_hairpin_unbind;
-   rte_eth_link_speed_to_str;
-   rte_eth_link_to_str;
-   rte_eth_fec_get_capability;
-   rte_eth_fec_get;
-   rte_eth_fec_set;
-   rte_flow_tunnel_decap_set;
-   rte_flow_tunnel_match;
-   rte_flow_get_restore_info;
-   rte_flow_tunnel_action_decap_release;
-   rte_flow_tunnel_item_release;
-
# added in 21.02
rte_eth_get_monitor_addr;
 
-- 
2.41.0

[PATCH v2 2/2] devtools: list symbols by version

[David Marchand]

Make it easier to list experimental symbols added in a certain version.

Example:
$ ./buildtools/map-list-symbol.sh -V 18.11 lib/eal/version.map
lib/eal/version.map EXPERIMENTAL rte_dev_event_callback_process
lib/eal/version.map EXPERIMENTAL rte_dev_hotplug_handle_disable
lib/eal/version.map EXPERIMENTAL rte_dev_hotplug_handle_enable

Signed-off-by: David Marchand 
Acked-by: Stephen Hemminger 
---
 buildtools/map-list-symbol.sh | 39 ++-
 1 file changed, 25 insertions(+), 14 deletions(-)

diff --git a/buildtools/map-list-symbol.sh b/buildtools/map-list-symbol.sh
index 3bf9bd66f8..a834399816 100755
--- a/buildtools/map-list-symbol.sh
+++ b/buildtools/map-list-symbol.sh
@@ -6,7 +6,7 @@ section=all
 symbol=all
 quiet=
 
-while getopts 'S:s:q' name; do
+while getopts 'S:s:qV:' name; do
case $name in
S)
[ $section = 'all' ] || {
@@ -25,8 +25,11 @@ while getopts 'S:s:q' name; do
q)
quiet='y'
;;
+   V)
+   version=$OPTARG
+   ;;
?)
-   echo 'usage: $0 [-S section] [-s symbol] [-q]'
+   echo 'usage: $0 [-S section] [-s symbol] [-V version] [-q]'
exit 1
;;
esac
@@ -38,7 +41,8 @@ for file in $@; do
cat "$file" |awk '
BEGIN {
current_section = "";
-   if ("'$section'" == "all" && "'$symbol'" == "all") {
+   current_version = "";
+   if ("'$section'" == "all" && "'$symbol'" == "all" && 
"'$version'" == "") {
ret = 0;
} else {
ret = 1;
@@ -49,18 +53,25 @@ for file in $@; do
current_section = $1;
}
}
-   /.*}/ { current_section = ""; }
+   /.*}/ { current_section = ""; current_version = ""; }
+   /^\t# added in / {
+   current_version=$4;
+   }
/^[^}].*[^:*];/ {
-   if (current_section != "") {
-   gsub(";","");
-   if ("'$symbol'" == "all" || $1 == "'$symbol'") {
-   ret = 0;
-   if ("'$quiet'" == "") {
-   print "'$file' "current_section" "$1;
-   }
-   if ("'$symbol'" != "all") {
-   exit 0;
-   }
+   if (current_section == "") {
+   next;
+   }
+   if ("'$version'" != "" && "'$version'" != current_version) {
+   next;
+   }
+   gsub(";","");
+   if ("'$symbol'" == "all" || $1 == "'$symbol'") {
+   ret = 0;
+   if ("'$quiet'" == "") {
+   print "'$file' "current_section" "$1;
+   }
+   if ("'$symbol'" != "all") {
+   exit 0;
}
}
}
-- 
2.41.0

Re: [PATCH v5 1/8] net/rnp: add skeleton

[Thomas Monjalon]

Hi,

Wenbo Cao:
> --- /dev/null
> +++ b/doc/guides/nics/rnp.rst
> @@ -0,0 +1,43 @@
> +..  SPDX-License-Identifier: BSD-3-Clause
> +Copyright(c) 2023 Mucse IC Design Ltd.
> +
> +RNP Poll Mode driver
> +==

Please keep underlining the same size as the text above.

> +
> +The RNP ETHDEV PMD (**librte_net_rnp**) provides poll mode ethdev
> +driver support for the inbuilt network device found in the **Mucse RNP**
> +
> +Prerequisites
> +-
> +More information can be found at `Mucse, Official Website
> +`_.
> +
> +Supported RNP SoCs
> +
> +
> +- N10
> +
> +Driver compilation and testing
> +--
> +
> +Refer to the document :ref:`compiling and testing a PMD for a NIC 
> `
> +for details.

It was a mistake to originally introduce the anchor "pmd_build_and_test".
You should achieve the same result with the shorter syntax :doc:`build_and_test`

> +
> +#. Running testpmd:
> +
> +   Follow instructions available in the document
> +   :ref:`compiling and testing a PMD for a NIC `
> +   to run testpmd.

Do we really need that referencing the same document as above?

> +
> +Limitations or Known issues
> +
> +Build with ICC is not supported yet.
> +CRC stripping
> +~~
> +The RNP SoC family NICs strip the CRC for every packets coming into the
> +host interface irrespective of the offload configuration.
> +When You Want To Disable CRC_OFFLOAD The Feature Will Influence The RxCksum 
> Offload
> +VLAN Strip
> +~~~
> +For VLAN Strip RNP Just Support CVLAN(0x8100) Type If The Vlan Type Is 
> SVLAN(0X88a8)
> +VLAN Filter Or Strip Will Not Effert For This Packet It Will Bypass To The 
> Host.

Please check the doc contribution guide.
You should add spaces before and after titles.

[PATCH v1 3/3] net/hns3: report maximum buffer size

[Huisong Li]

This patch reports the maximum buffer size hardware supported.

Signed-off-by: Huisong Li 
---
 drivers/net/hns3/hns3_common.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/hns3/hns3_common.c b/drivers/net/hns3/hns3_common.c
index a11ea686fd..6a34df2d75 100644
--- a/drivers/net/hns3/hns3_common.c
+++ b/drivers/net/hns3/hns3_common.c
@@ -59,6 +59,7 @@ hns3_dev_infos_get(struct rte_eth_dev *eth_dev, struct 
rte_eth_dev_info *info)
info->max_tx_queues = hw->tqps_num;
info->max_rx_pktlen = HNS3_MAX_FRAME_LEN; /* CRC included */
info->min_rx_bufsize = HNS3_MIN_BD_BUF_SIZE;
+   info->max_rx_bufsize = HNS3_MAX_BD_BUF_SIZE;
info->max_mtu = info->max_rx_pktlen - HNS3_ETH_OVERHEAD;
info->max_lro_pkt_size = HNS3_MAX_LRO_SIZE;
info->rx_offload_capa = (RTE_ETH_RX_OFFLOAD_IPV4_CKSUM |
-- 
2.33.0

[PATCH v1 2/3] app/testpmd: add maximum Rx buffer size display

[Huisong Li]

Add maximum Rx buffer size display.

Signed-off-by: Huisong Li 
---
 app/test-pmd/config.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 11f3a22048..e0a11a35fe 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -880,6 +880,8 @@ port_infos_display(portid_t port_id)
}
 
printf("Minimum size of RX buffer: %u\n", dev_info.min_rx_bufsize);
+   if (dev_info.max_rx_bufsize != UINT32_MAX)
+   printf("Maximum size of RX buffer: %u\n", 
dev_info.max_rx_bufsize);
printf("Maximum configurable length of RX packet: %u\n",
dev_info.max_rx_pktlen);
printf("Maximum configurable size of LRO aggregated packet: %u\n",
-- 
2.33.0

[PATCH v1 0/3] introduce maximum Rx buffer size

[Huisong Li]

The Rx buffer size stands for the size hardware supported to receive
packets in one mbuf. The "min_rx_bufsize" is the minimum buffer hardware
supported in Rx. Actually, network engines also have the maximum buffer.
So this series introduce maximum Rx buffer size in struct rte_eth_dev_info
and display it by testpmd.

---
v1:
 - move max_rx_bufsize to min_rx_bufsize closer in struct rte_eth_dev_info
 - add max_rx_bufsize display in testpmd.
 - hns3 reports maximum buffer size.

Huisong Li (3):
  ethdev: introduce maximum Rx buffer size
  app/testpmd: add maximum Rx buffer size display
  net/hns3: report maximum buffer size

 app/test-pmd/config.c  | 2 ++
 drivers/net/hns3/hns3_common.c | 1 +
 lib/ethdev/rte_ethdev.c| 7 +++
 lib/ethdev/rte_ethdev.h| 6 ++
 4 files changed, 16 insertions(+)

-- 
2.33.0

[PATCH v1 1/3] ethdev: introduce maximum Rx buffer size

[Huisong Li]

The Rx buffer size stands for the size hardware supported to receive
packets in one mbuf. The "min_rx_bufsize" is the minimum buffer hardware
supported in Rx. Actually, some engines also have the maximum buffer
specification, like, hns3. For these engines, the available data size
of one mbuf in Rx also depends on the maximum buffer hardware supported.
So introduce maximum Rx buffer size in struct rte_eth_dev_info to report
user to avoid memory waste. And driver should accept it and just pass
maximum buffer hardware supported to hardware if application specifies
the Rx buffer size is greater than the maximum buffer.

Signed-off-by: Huisong Li 
---
 lib/ethdev/rte_ethdev.c | 7 +++
 lib/ethdev/rte_ethdev.h | 6 ++
 2 files changed, 13 insertions(+)

diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index 0840d2b594..9985bd3049 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -2068,6 +2068,7 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t 
rx_queue_id,
struct rte_eth_dev *dev;
struct rte_eth_dev_info dev_info;
struct rte_eth_rxconf local_conf;
+   uint32_t vld_bufsize;
 
RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
dev = &rte_eth_devices[port_id];
@@ -2105,6 +2106,11 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t 
rx_queue_id,
return ret;
 
mbp_buf_size = rte_pktmbuf_data_room_size(mp);
+   vld_bufsize = mbp_buf_size - RTE_PKTMBUF_HEADROOM;
+   if (vld_bufsize > dev_info.max_rx_bufsize)
+   RTE_ETHDEV_LOG(WARNING,
+   "Ethdev port_id=%u Rx buffer size (%u) is 
greater than the maximum buffer size (%u) driver supported.\n",
+   port_id, vld_bufsize, dev_info.max_rx_bufsize);
} else if (rx_conf != NULL && rx_conf->rx_nseg > 0) {
const struct rte_eth_rxseg_split *rx_seg;
uint16_t n_seg;
@@ -3689,6 +3695,7 @@ rte_eth_dev_info_get(uint16_t port_id, struct 
rte_eth_dev_info *dev_info)
dev_info->min_mtu = RTE_ETHER_MIN_LEN - RTE_ETHER_HDR_LEN -
RTE_ETHER_CRC_LEN;
dev_info->max_mtu = UINT16_MAX;
+   dev_info->max_rx_bufsize = UINT32_MAX;
 
if (*dev->dev_ops->dev_infos_get == NULL)
return -ENOTSUP;
diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
index 04a2564f22..9fdf2c75ee 100644
--- a/lib/ethdev/rte_ethdev.h
+++ b/lib/ethdev/rte_ethdev.h
@@ -1724,6 +1724,12 @@ struct rte_eth_dev_info {
uint16_t max_mtu;   /**< Maximum MTU allowed */
const uint32_t *dev_flags; /**< Device flags */
uint32_t min_rx_bufsize; /**< Minimum size of Rx buffer. */
+   /**
+* Maximum size of Rx buffer. Driver should accept it and just pass
+* this value to HW if application specifies the Rx buffer size is
+* greater than this value.
+*/
+   uint32_t max_rx_bufsize;
uint32_t max_rx_pktlen; /**< Maximum configurable length of Rx pkt. */
/** Maximum configurable size of LRO aggregated packet. */
uint32_t max_lro_pkt_size;
-- 
2.33.0

NVIDIA roadmap for 23.11

[Maayan Kashani]

Please find below NVIDIA roadmap for 23.11 release:

A. rte_flow new APIs
=

  1.  Support color-aware mode for meters in shared actions list
Support meter color updates based on traffic BW. Additionally, enable use of 
shared (indirect) meters as part of shared action lists


  1.  Add IPv6 extension push & remove actions.
Add new rte_flow actions to support push/remove IPv6 extension header.​


  1.  Add template API new rte_flow function for hash calculation to enable SW 
applications to use the same hash as calculated in HW.
This capabiity enables alignment of hash values between HW pipelines and SW 
pipelines, so partial offloads pipes can be aligned with offloaded pipes and 
use the same hash value.
When an application uses hash-based insert by index, the flow of the packet 
depends on the calculated hash. For packets that belong to the same flow
but passed through the SW such hash was not calculated. In this case, 
application can calculate the same hash as the HW does before sending it to the 
wire.​


  1.  Support group default missaction

A group's miss actions are a set of actions to be performed in case of a miss 
on a group, meaning a packet didn't hit any rules in the group.

This new API function allows a user to set a group's miss actions in an 
explicit way, reducing number of hops and increasing pps in case of handling 
misses.


  1.  Add RTE_FLOW_ITEM_TYPE_PTYPE to allow matching on L2/L3/L4 and tunnel 
information as defined in mbuf.
The packet type matching provides quick way of finding out L2/L3/L4 protocols 
in each packet.
That helps with optimized flow rules matching, eliminating the need of stacking 
all the packet headers in the matching criteria.
As part of this feature a new packet type will be added for ESP.


  1.  Adding NAT64 action
Adding a new action for header re-write of NAT64.
NAT64 (Network Address Translation 64) is a networking technology that 
facilitates communication between IPv6 and IPv4 networks.


B. Net/mlx5 PMD updates
=
All new APIs implemented above will be supported in mlx5.
Other features described below:


  1.  Provide additional 3 tags in the new generation NICs with rte_flow 
template API​.
With template API, 3 tags are available in extended meta mode (dv_xmeta_en=4) 
and 5 tags are available in legacy mode for offloading. Starting with 
ConnectX-7 / BlueField-3,
additional 3 tags are exposed and can be used for matching and modification as 
the other tags. By incorporating additional tags, the application gains greater 
flexibility to
accommodate a wide range of use cases by enriching the packets with contextual 
data. This enables better abstraction of applications and streamlines the 
pipelines.


  1.  Support parsing of VXLAN-gpe with NSH
VXLAN-gpe with variable Network Service Headers (NSH) offers greater 
flexibility compared to certain other protocols, such as GRE.​
The presence of an NSH header can be checked and matched, without matching any 
specific field of the NSH. Then the inner header can also be matched and used 
as RSS fields.
The VXLAN-gpe with a variable NSH can be parsed and recognized by the NIC. In 
the rte_flow non-template API, the presence of NSH can be matched.
The inner header after VXLAN-gpe and NSH can be used for matching and for RSS 
hash fields.​
In the rte_flow non-template API, The VXLAN-gpe with a variable NSH can be 
parsed and recognized by the NIC. The presence of NSH header can be
matched and the inner header after VXLAN-gpe and NSH can be used for matching 
and for RSS hash fields.​


  1.  Support Geneve encap/decap and modify actions.
Added support for Geneve raw_encap/decap as well as Geneve options modify with 
template API in addition to previously supported match on Geneve options.​


  1.  Extend mirror action to support mirroring from FDB to queue/RSS.
Enable packets steering from FDB domain to SW through supporting explicit 
PORT_REPRESENTOR action for steering to SW. Useful for mirroring use case in 
which mirrored packets need to be destined to the SW for diagnostic.​


  1.  Support sharing raw_encap/decap action with indirect action handle.
In some cases when using template API, large number of rules have the same 
encap/decap actions characteristics. To reduce the number of encap/decap 
instances, we provide a method to share the
encap/decap actions amongst the different rules through definition of these 
shared encap/decap actions as indirect (stand-alone) entities that are reused 
and shared by multiple rules.
This indirect encap/decap action will be shared by different flow rules among 
different tables. ​
Using this method, memory footprint and cache-misses would be reduced, 
increasing overall PPS and enabling use of higher scale of flows through 
optimized resources. ​


  1.  Reuse encap/modify actions for different templates.
Allows for a reduction in memory footprint when *multiple* action *templates* 
with encap/modify actions are shared for the sam

Re: [PATCH] dma/skeleton: fix pthread set affinity

[David Marchand]

On Mon, Aug 14, 2023 at 6:24 PM Tyler Retzlaff
 wrote:
>
> On Sat, Aug 12, 2023 at 07:27:38AM +0530, Vipin Varghese wrote:
> > In skeleton dma driver, the user can pin the dma thread to desired
> > cpu core. The previous commit changed the api from set-affinity to
> > get-affinity leading to thread to be pinned to first available lcore.
> >
> > Bugzilla ID: 1270
> > Fixes: 7fdf30d8b259 ("eal/windows: remove most pthread lifetime shim 
> > functions")

This Fixes: looks wrong.

I would blame (and blame the bad reviewers):
Fixes: f241553b43ee ("dma/skeleton: drop some pthread API calls")

Can you double check?


-- 
David Marchand

[PATCH v4 0/8] expand list of optional libraries

[Bruce Richardson]

This patchset contains changes to DPDK to make the build more configurable
overall. The changes can be grouped into two areas:

* Firstly, there are changes to make the build of the unit tests more
  flexible and maintainable. These 5 patches switch the unit tests from
  having separate hard-coded lists of files and tests, with lots of
  conditional logic to add/remove items from the list, to instead being
  a single list of files and dependencies per-file. The actual lists of
  test-suites are built up dynamically from the list of files whose
  dependencies are met and therefore will be part of the build.
  This rework enables...
* The second set of patches which expands the list of libraries which
  can be disabled at build-time, while still giving a successful build
  of DPDK. Overall, after this set is applied, the list of optional
  libraries in DPDK is approx doubled in size. [15 additions giving
  31 optional libs]

NOTES:
* as stated in the RFC, the test binary moves in the build as a result
  of this set - from app/test/dpdk-test to app/dpdk-test.
* running the suites of tests no longer works after patch 4 until patch
  5 is applied. If this is an issue, the two patches can be merged on
  apply. However, the individual tests are still runnable through the
  test binary - it's only through "meson test" that there are issues, so
  I'm keeping the patches separate since each is relatively complex in its
  own right.
* A number of files in the test directory were not being built, and seem
  to have never been built with meson. This requires separate work, outside
  the scope of the patchset to fix.
* Similarly, some unit tests were never added to test suites. These can
  now be found by using grep to find the old REGISTER_TEST_COMMAND macro.
  Again, adding orphan tests to suites is outside the scope of this set.

V4:
- improve test suite definition, hopefully improving test pass rate:
  - drop unnecessary file-prefix for fast tests, since tests are not parallel
  - always pass driver path flag in "meson test" for shared builds.

V3:
- Create symlink to old path to test binary, to ensure all automated CI,
  and any other use of the test binary still works.

V2:
- Changes since RFC:
   -  merged two previous sets - reworking test build, and disabling
  libs - into one, since they depend on each other
   -  Reworked the list of dependencies for each file in the tests,
  so that files that depend on other files have those deps called
  out.
   -  Added list of external dependencies to the test builds
   -  For disabling libs, merged patches disabling individual libs
  together into one patch, since all unit test build changes were
  unnecessary. This made each patch a single-line addition to
  the optional list in lib/meson.build.

Bruce Richardson (8):
  app/test: add new macros for various test types
  app/test: tag tests with the test type
  app/test: make telemetry data test buildable on windows
  app/test: build using per-file dependency matrix
  app/test: define unit tests suites based on test macros
  examples/l3fwd: make eventdev an optional dependency
  build: make most device classes optional
  build: expand list of optional libraries

 app/meson.build   |  15 +-
 app/test/meson.build  | 823 +-
 app/test/suites/meson.build   |  72 ++
 app/test/test.h   |   9 +-
 app/test/test_acl.c   |   2 +-
 app/test/test_atomic.c|   2 +-
 app/test/test_barrier.c   |   2 +-
 app/test/test_bitmap.c|   2 +-
 app/test/test_bitops.c|   2 +-
 app/test/test_bpf.c   |   4 +-
 app/test/test_byteorder.c |   2 +-
 app/test/test_cksum.c |   2 +-
 app/test/test_cmdline.c   |   2 +-
 app/test/test_common.c|   2 +-
 app/test/test_cpuflags.c  |   2 +-
 app/test/test_crc.c   |   2 +-
 app/test/test_cryptodev.c |  38 +-
 app/test/test_cryptodev_asym.c|   2 +-
 app/test/test_cycles.c|   2 +-
 app/test/test_debug.c |   2 +-
 app/test/test_devargs.c   |   2 +-
 app/test/test_distributor.c   |   2 +-
 app/test/test_distributor_perf.c  |   2 +-
 app/test/test_dmadev.c|   2 +-
 app/test/test_eal_flags.c |  24 +-
 app/test/test_eal_fs.c|   2 +-
 app/test/test_efd.c   |   2 +-
 app/test/test_efd_perf.c  |   2 +-
 app/test/test_errno.c |   2 +-
 app/test/test_ethdev_link.c   |   2 +-
 app/test/test_event_ring.c|   2 +-
 app/test/test_ev

[PATCH v4 1/8] app/test: add new macros for various test types

[Bruce Richardson]

Rather than just registering all tests using a single generic macro,
add macros which identify the test as being of a particular type.

Signed-off-by: Bruce Richardson 
---
 app/test/test.h | 9 -
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/app/test/test.h b/app/test/test.h
index 85f57efbc6..a91ded76af 100644
--- a/app/test/test.h
+++ b/app/test/test.h
@@ -190,7 +190,7 @@ struct test_command {
 
 void add_test_command(struct test_command *t);
 
-/* Register a test function with its command string */
+/* Register a test function with its command string. Should not be used 
directly */
 #define REGISTER_TEST_COMMAND(cmd, func) \
static struct test_command test_struct_##cmd = { \
.command = RTE_STR(cmd), \
@@ -201,4 +201,11 @@ void add_test_command(struct test_command *t);
add_test_command(&test_struct_##cmd); \
}
 
+/* Register a test function as a particular type.
+ * These can be used to build up test suites automatically
+ */
+#define REGISTER_FAST_TEST(cmd, no_huge, ASan, func)  
REGISTER_TEST_COMMAND(cmd, func)
+#define REGISTER_PERF_TEST REGISTER_TEST_COMMAND
+#define REGISTER_DRIVER_TEST REGISTER_TEST_COMMAND
+
 #endif
-- 
2.39.2

[PATCH v4 2/8] app/test: tag tests with the test type

[Bruce Richardson]

Rather than having the test types called out in the meson.build file, we
can use macros to identify the test type in the C file itself and then
dynamically build up the tests lists at config time.

Signed-off-by: Bruce Richardson 
---
 app/test/test_acl.c  |  2 +-
 app/test/test_atomic.c   |  2 +-
 app/test/test_barrier.c  |  2 +-
 app/test/test_bitmap.c   |  2 +-
 app/test/test_bitops.c   |  2 +-
 app/test/test_bpf.c  |  4 +--
 app/test/test_byteorder.c|  2 +-
 app/test/test_cksum.c|  2 +-
 app/test/test_cmdline.c  |  2 +-
 app/test/test_common.c   |  2 +-
 app/test/test_cpuflags.c |  2 +-
 app/test/test_crc.c  |  2 +-
 app/test/test_cryptodev.c| 38 
 app/test/test_cryptodev_asym.c   |  2 +-
 app/test/test_cycles.c   |  2 +-
 app/test/test_debug.c|  2 +-
 app/test/test_devargs.c  |  2 +-
 app/test/test_distributor.c  |  2 +-
 app/test/test_distributor_perf.c |  2 +-
 app/test/test_dmadev.c   |  2 +-
 app/test/test_eal_flags.c| 24 +++
 app/test/test_eal_fs.c   |  2 +-
 app/test/test_efd.c  |  2 +-
 app/test/test_efd_perf.c |  2 +-
 app/test/test_errno.c|  2 +-
 app/test/test_ethdev_link.c  |  2 +-
 app/test/test_event_ring.c   |  2 +-
 app/test/test_eventdev.c |  2 +-
 app/test/test_fbarray.c  |  2 +-
 app/test/test_fib.c  |  4 +--
 app/test/test_fib6.c |  4 +--
 app/test/test_fib6_perf.c|  2 +-
 app/test/test_fib_perf.c |  2 +-
 app/test/test_func_reentrancy.c  |  2 +-
 app/test/test_hash.c |  2 +-
 app/test/test_hash_functions.c   |  2 +-
 app/test/test_hash_multiwriter.c |  2 +-
 app/test/test_hash_perf.c|  2 +-
 app/test/test_hash_readwrite.c   |  4 +--
 app/test/test_hash_readwrite_lf_perf.c   |  2 +-
 app/test/test_interrupts.c   |  2 +-
 app/test/test_ipfrag.c   |  2 +-
 app/test/test_ipsec.c|  2 +-
 app/test/test_ipsec_perf.c   |  2 +-
 app/test/test_kvargs.c   |  2 +-
 app/test/test_lcores.c   |  2 +-
 app/test/test_logs.c |  2 +-
 app/test/test_lpm.c  |  2 +-
 app/test/test_lpm6.c |  2 +-
 app/test/test_lpm6_perf.c|  2 +-
 app/test/test_lpm_perf.c |  2 +-
 app/test/test_malloc.c   |  2 +-
 app/test/test_malloc_perf.c  |  2 +-
 app/test/test_mbuf.c |  2 +-
 app/test/test_mcslock.c  |  2 +-
 app/test/test_member.c   |  2 +-
 app/test/test_member_perf.c  |  2 +-
 app/test/test_memcpy.c   |  2 +-
 app/test/test_memcpy_perf.c  |  2 +-
 app/test/test_memory.c   |  2 +-
 app/test/test_mempool.c  |  2 +-
 app/test/test_mempool_perf.c |  2 +-
 app/test/test_memzone.c  |  2 +-
 app/test/test_meter.c|  2 +-
 app/test/test_mp_secondary.c |  2 +-
 app/test/test_per_lcore.c|  2 +-
 app/test/test_pflock.c   |  2 +-
 app/test/test_pie.c  |  6 ++--
 app/test/test_pmd_perf.c |  2 +-
 app/test/test_power.c|  2 +-
 app/test/test_power_cpufreq.c|  2 +-
 app/test/test_power_intel_uncore.c   |  2 +-
 app/test/test_power_kvm_vm.c |  2 +-
 app/test/test_prefetch.c |  2 +-
 app/test/test_rand_perf.c|  2 +-
 app/test/test_rcu_qsbr.c |  2 +-
 app/test/test_rcu_qsbr_perf.c|  2 +-
 app/test/test_reassembly_perf.c  |  2 +-
 app/test/test_reciprocal_division.c  |  2 +-
 app/test/test_reciprocal_division_perf.c |  2 +-
 app/test/test_red.c  |  4 +--
 app/test/test_reorder.c  |  2 +-
 app/test/test_rib.c  |  4 +--
 app/test/test_rib6.c |  4 +--
 app/test/test_ring.c |  2 +-
 app/test/test_ring_perf.c|  2 +-
 app/test/test_rwlock.c   |  8 ++---
 app/test/test_sched.c|  2 +-
 app/test/test_security.c |  2 +-
 app/test/test_seqlock.c  |  2 +-
 app/test/test_service_cores.c|  4 +--
 app/test/test_spinlock.c |  2 +-
 app/test/test_stack.c|  4 +--
 app/test/test_stack_perf.c   |  4 +--
 app/test/test

[PATCH v4 3/8] app/test: make telemetry data test buildable on windows

[Bruce Richardson]

To enable the building of the telemetry data tests file when building on
windows, we need to provide a stub implementation. That way, the test
file is buildable any time the library itself is built.

Signed-off-by: Bruce Richardson 
---
 app/test/test_telemetry_data.c | 14 --
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/app/test/test_telemetry_data.c b/app/test/test_telemetry_data.c
index a960f44c1b..b2dc1d75db 100644
--- a/app/test/test_telemetry_data.c
+++ b/app/test/test_telemetry_data.c
@@ -2,11 +2,20 @@
  * Copyright 2020 Intel Corporation
  */
 
+#ifdef RTE_EXEC_ENV_WINDOWS
+#include "test.h"
+
+static int
+telemetry_data_autotest(void)
+{
+   return TEST_SKIPPED;
+}
+
+#else
+
 #include 
 #include 
-#ifndef RTE_EXEC_ENV_WINDOWS
 #include 
-#endif
 #include 
 #include 
 
@@ -604,5 +613,6 @@ telemetry_data_autotest(void)
close(sock);
return 0;
 }
+#endif /* windows/non-windows */
 
 REGISTER_TEST_COMMAND(telemetry_data_autotest, telemetry_data_autotest);
-- 
2.39.2

[PATCH v4 4/8] app/test: build using per-file dependency matrix

[Bruce Richardson]

Rather than using if-else constructs to selectively add or remove files
from the UT build, switch to a table-based approach where each file
lists out what libs or drivers it depends upon.

Initial version of this table was generated via analysis of the header
files included in each C file. The basic dependencies of the test binary
[cmdline, ring, mempool and mbuf] were then removed from the per-file
lists, as there is no point in checking them as the whole app will be
disabled if they are not present.

With the file list, the dependencies for the "utility" C-files are kept
separate, so that other tests which depend on the functions provided by
those files can have that dependency recorded properly. The basic
cryptodev tests also fall into this category as functions from the main
cryptodev test file are used by other crypto tests.

As well as the main table for internal dependencies, some test
files have separate external components too. A second, much smaller
table lists these dependencies.

Signed-off-by: Bruce Richardson 
---
 app/meson.build  |   8 +-
 app/test/meson.build | 823 ---
 2 files changed, 238 insertions(+), 593 deletions(-)

diff --git a/app/meson.build b/app/meson.build
index 4fc1a83eba..0d8b618e7f 100644
--- a/app/meson.build
+++ b/app/meson.build
@@ -32,6 +32,11 @@ apps = [
 'test-security-perf',
 ]
 
+if get_option('tests')
+# build the auto test app if enabled.
+apps += 'test'
+endif
+
 default_cflags = machine_args + ['-DALLOW_EXPERIMENTAL_API']
 default_ldflags = []
 if get_option('default_library') == 'static' and not is_windows
@@ -106,6 +111,3 @@ foreach app:apps
 install_rpath: join_paths(get_option('prefix'), 
driver_install_path),
 install: true)
 endforeach
-
-# special case the autotests
-subdir('test')
diff --git a/app/test/meson.build b/app/test/meson.build
index 66897c14a3..2b885ae273 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -1,446 +1,230 @@
 # SPDX-License-Identifier: BSD-3-Clause
-# Copyright(c) 2017 Intel Corporation
-
-if not get_option('tests')
-subdir_done()
-endif
-
-test_sources = files(
-'commands.c',
-'packet_burst_generator.c',
-'test.c',
-'test_acl.c',
-'test_alarm.c',
-'test_atomic.c',
-'test_barrier.c',
-'test_bitops.c',
-'test_bitmap.c',
-'test_bpf.c',
-'test_byteorder.c',
-'test_cksum.c',
-'test_cksum_perf.c',
-'test_cmdline.c',
-'test_cmdline_cirbuf.c',
-'test_cmdline_etheraddr.c',
-'test_cmdline_ipaddr.c',
-'test_cmdline_lib.c',
-'test_cmdline_num.c',
-'test_cmdline_portlist.c',
-'test_cmdline_string.c',
-'test_common.c',
-'test_cpuflags.c',
-'test_crc.c',
-'test_cryptodev.c',
-'test_cryptodev_asym.c',
-'test_cryptodev_blockcipher.c',
-'test_cryptodev_crosscheck.c',
-'test_cryptodev_security_ipsec.c',
-'test_cryptodev_security_pdcp.c',
-'test_cycles.c',
-'test_debug.c',
-'test_devargs.c',
-'test_distributor.c',
-'test_distributor_perf.c',
-'test_dmadev.c',
-'test_dmadev_api.c',
-'test_eal_flags.c',
-'test_eal_fs.c',
-'test_efd.c',
-'test_efd_perf.c',
-'test_errno.c',
-'test_ethdev_link.c',
-'test_event_crypto_adapter.c',
-'test_event_eth_rx_adapter.c',
-'test_event_ring.c',
-'test_event_timer_adapter.c',
-'test_eventdev.c',
-'test_external_mem.c',
-'test_fbarray.c',
-'test_fib.c',
-'test_fib_perf.c',
-'test_fib6.c',
-'test_fib6_perf.c',
-'test_func_reentrancy.c',
-'test_hash.c',
-'test_hash_functions.c',
-'test_hash_multiwriter.c',
-'test_hash_readwrite.c',
-'test_hash_perf.c',
-'test_hash_readwrite_lf_perf.c',
-'test_interrupts.c',
-'test_ipfrag.c',
-'test_ipsec.c',
-'test_ipsec_sad.c',
-'test_ipsec_perf.c',
-'test_kvargs.c',
-'test_lcores.c',
-'test_logs.c',
-'test_lpm.c',
-'test_lpm6.c',
-'test_lpm6_perf.c',
-'test_lpm_perf.c',
-'test_malloc.c',
-'test_malloc_perf.c',
-'test_mbuf.c',
-'test_member.c',
-'test_member_perf.c',
-'test_memcpy.c',
-'test_memcpy_perf.c',
-'test_memory.c',
-'test_mempool.c',
-'test_mempool_perf.c',
-'test_memzone.c',
-'test_meter.c',
-'test_mcslock.c',
-'test_mp_secondary.c',
-'test_per_lcore.c',
-'test_pflock.c',
-'test_pmd_perf.c',
-'test_power.c',
-'test_power_cpufreq.c',
-'test_power_kvm_vm.c',
-'test_power_intel_uncore.c',
-'test_prefetch.c',
-'test_rand_perf.c',
-

[PATCH v4 5/8] app/test: define unit tests suites based on test macros

[Bruce Richardson]

Rather than having the test suites listed out in the meson.build files
and having to have them enabled/disabled selectively based on what libs
are being built, pull the tests to run from the source files which were
added to the build.

Most test suites require no additional info other than the list of test
names in the suite. However the fast-test are special that they have
additional parameters associated with them. This requires some
additional work in the test extraction script and in the meson.build
file for processing the output.

Signed-off-by: Bruce Richardson 
---
 app/meson.build   |  7 +-
 app/test/suites/meson.build   | 72 +++
 buildtools/get-test-suites.py | 33 +
 .../has-hugepages.py  |  0
 buildtools/meson.build|  2 +
 5 files changed, 113 insertions(+), 1 deletion(-)
 create mode 100644 app/test/suites/meson.build
 create mode 100644 buildtools/get-test-suites.py
 rename app/test/has_hugepage.py => buildtools/has-hugepages.py (100%)

diff --git a/app/meson.build b/app/meson.build
index 0d8b618e7f..c14dc80892 100644
--- a/app/meson.build
+++ b/app/meson.build
@@ -101,7 +101,7 @@ foreach app:apps
 link_libs = dpdk_static_libraries + dpdk_drivers
 endif
 
-executable('dpdk-' + name,
+exec = executable('dpdk-' + name,
 sources,
 c_args: cflags,
 link_args: ldflags,
@@ -110,4 +110,9 @@ foreach app:apps
 include_directories: includes,
 install_rpath: join_paths(get_option('prefix'), 
driver_install_path),
 install: true)
+if name == 'test'
+dpdk_test = exec
+autotest_sources = sources
+subdir('test/suites')  # define the pre-canned test suites
+endif
 endforeach
diff --git a/app/test/suites/meson.build b/app/test/suites/meson.build
new file mode 100644
index 00..86884480a4
--- /dev/null
+++ b/app/test/suites/meson.build
@@ -0,0 +1,72 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2023 Intel Corporation
+
+# some perf tests (eg: memcpy perf autotest)take very long
+# to complete, so timeout to 10 minutes
+timeout_seconds = 600
+timeout_seconds_fast = 10
+
+test_no_huge_args = ['--no-huge', '-m', '2048']
+has_hugepage = run_command(has_hugepages_cmd, check: true).stdout().strip() != 
'0'
+message('hugepage availability: @0@'.format(has_hugepage))
+
+# process source files to determine the different unit test suites
+# - fast_tests
+# - perf_tests
+# - driver_tests
+test_suites = run_command(get_test_suites_cmd, autotest_sources,
+ check: true).stdout().strip().split()
+foreach suite:test_suites
+# simple cases - tests without parameters or special handling
+suite = suite.split('=')
+suite_name = suite[0]
+suite_tests = suite[1].split(',')
+if suite_name != 'fast-tests'
+# simple cases - tests without parameters or special handling
+foreach t: suite_tests
+test(t, dpdk_test,
+args: ['--', t],
+timeout: timeout_seconds,
+is_parallel: false,
+suite: suite_name)
+endforeach
+else
+# special fast-test handling here
+foreach t: suite_tests
+params = t.split(':')
+test_name = params[0]
+nohuge = params[1] == 'true'
+asan = params[2] == 'true'
+
+test_args = []
+if nohuge
+test_args += test_no_huge_args
+elif not has_hugepage
+continue  #skip this tests
+endif
+if not asan and (get_option('b_sanitize') == 'address'
+or get_option('b_sanitize') == 'address,undefined')
+continue  # skip this test
+endif
+
+if get_option('default_library') == 'shared'
+test_args += ['-d', dpdk_drivers_build_dir]
+endif
+
+test(test_name, dpdk_test,
+args : test_args + ['--', test_name],
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'fast-tests')
+if not is_windows and test_name == 'trace_autotest'
+test_args += ['--trace=.*']
+test_args += 
['--trace-dir=@0@'.format(meson.current_build_dir())]
+test(test_name + '_with_traces', dpdk_test,
+args : test_args + ['--', test_name],
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'fast-tests')
+endif
+endforeach
+endif
+endforeach
diff --git a/buildtools/get-test-suites.py b/buildtools/get-test-suites.py
new file mode 100644
index 00..95a9cad4c8
--- /dev/null
+++ b/buildtools/get-test-suites.py
@@ -0,0 +1,33 @@
+#! /usr/bin/env python3
+# SPDX-License-Ide

[PATCH v4 6/8] examples/l3fwd: make eventdev an optional dependency

[Bruce Richardson]

With l3fwd being a very commonly used example app, and built as part of
our CI tests, we need to ensure it's buildable with just about all
supported DPDK configurations.  To enable l3fwd application to be built
when the eventdev library is disabled, we need to compile in the
eventdev support conditionally.  Thankfully, the eventdev support is
pretty self-contained, with only the main.c file having more than a
couple of ifdefs.

Signed-off-by: Bruce Richardson 
---
 examples/l3fwd/l3fwd_em.c  |  2 +
 examples/l3fwd/l3fwd_em_hlm.h  |  2 +
 examples/l3fwd/l3fwd_event.c   |  2 +
 examples/l3fwd/l3fwd_event.h   |  7 ++-
 examples/l3fwd/l3fwd_event_generic.c   |  2 +
 examples/l3fwd/l3fwd_event_internal_port.c |  2 +
 examples/l3fwd/l3fwd_fib.c |  2 +
 examples/l3fwd/l3fwd_lpm.c |  2 +
 examples/l3fwd/main.c  | 65 +-
 examples/l3fwd/meson.build |  5 +-
 10 files changed, 74 insertions(+), 17 deletions(-)

diff --git a/examples/l3fwd/l3fwd_em.c b/examples/l3fwd/l3fwd_em.c
index 476ac0c54f..40e102b38a 100644
--- a/examples/l3fwd/l3fwd_em.c
+++ b/examples/l3fwd/l3fwd_em.c
@@ -663,6 +663,7 @@ em_main_loop(__rte_unused void *dummy)
return 0;
 }
 
+#ifdef RTE_LIB_EVENTDEV
 static __rte_always_inline void
 em_event_loop_single(struct l3fwd_event_resources *evt_rsrc,
const uint8_t flags)
@@ -959,6 +960,7 @@ em_event_main_loop_tx_q_burst_vector(__rte_unused void 
*dummy)
em_event_loop_vector(evt_rsrc, L3FWD_EVENT_TX_ENQ);
return 0;
 }
+#endif
 
 /* Initialize exact match (hash) parameters. 8< */
 void
diff --git a/examples/l3fwd/l3fwd_em_hlm.h b/examples/l3fwd/l3fwd_em_hlm.h
index 2e11eefad7..31cda9ddc1 100644
--- a/examples/l3fwd/l3fwd_em_hlm.h
+++ b/examples/l3fwd/l3fwd_em_hlm.h
@@ -255,6 +255,7 @@ l3fwd_em_send_packets(int nb_rx, struct rte_mbuf 
**pkts_burst, uint16_t portid,
send_packets_multi(qconf, pkts_burst, dst_port, nb_rx);
 }
 
+#ifdef RTE_LIB_EVENTDEV
 /*
  * Buffer optimized handling of events, invoked
  * from main_loop.
@@ -347,5 +348,6 @@ l3fwd_em_process_event_vector(struct rte_event_vector *vec,
 
process_event_vector(vec, dst_port);
 }
+#endif /* RTE_LIB_EVENTDEV */
 
 #endif /* __L3FWD_EM_HLM_H__ */
diff --git a/examples/l3fwd/l3fwd_event.c b/examples/l3fwd/l3fwd_event.c
index 32906ab08d..d72a4138cc 100644
--- a/examples/l3fwd/l3fwd_event.c
+++ b/examples/l3fwd/l3fwd_event.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 #include 
 
@@ -341,3 +342,4 @@ l3fwd_event_worker_cleanup(uint8_t event_d_id, uint8_t 
event_p_id,
rte_event_port_quiesce(event_d_id, event_p_id, l3fwd_event_port_flush,
   NULL);
 }
+#endif /* #ifdef RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_event.h b/examples/l3fwd/l3fwd_event.h
index e21817c36b..1fd6fe4a78 100644
--- a/examples/l3fwd/l3fwd_event.h
+++ b/examples/l3fwd/l3fwd_event.h
@@ -6,11 +6,13 @@
 #define __L3FWD_EVENTDEV_H__
 
 #include 
+#include 
+#include 
+
+#ifdef RTE_LIB_EVENTDEV
 #include 
 #include 
 #include 
-#include 
-#include 
 
 #include "l3fwd.h"
 
@@ -164,4 +166,5 @@ void l3fwd_event_worker_cleanup(uint8_t event_d_id, uint8_t 
event_p_id,
struct rte_event events[], uint16_t nb_enq,
uint16_t nb_deq, uint8_t is_vector);
 
+#endif /* ifdef RTE_LIB_EVENTDEV */
 #endif /* __L3FWD_EVENTDEV_H__ */
diff --git a/examples/l3fwd/l3fwd_event_generic.c 
b/examples/l3fwd/l3fwd_event_generic.c
index c80573fc58..ddb6e5c38d 100644
--- a/examples/l3fwd/l3fwd_event_generic.c
+++ b/examples/l3fwd/l3fwd_event_generic.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 
 #include "l3fwd.h"
@@ -309,3 +310,4 @@ l3fwd_event_set_generic_ops(struct l3fwd_event_setup_ops 
*ops)
ops->event_port_setup = l3fwd_event_port_setup_generic;
ops->adapter_setup = l3fwd_rx_tx_adapter_setup_generic;
 }
+#endif /* RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_event_internal_port.c 
b/examples/l3fwd/l3fwd_event_internal_port.c
index 32cf657148..cb49a8b9fa 100644
--- a/examples/l3fwd/l3fwd_event_internal_port.c
+++ b/examples/l3fwd/l3fwd_event_internal_port.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 
 #include "l3fwd.h"
@@ -311,3 +312,4 @@ l3fwd_event_set_internal_port_ops(struct 
l3fwd_event_setup_ops *ops)
ops->event_port_setup = l3fwd_event_port_setup_internal_port;
ops->adapter_setup = l3fwd_rx_tx_adapter_setup_internal_port;
 }
+#endif /* RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_fib.c b/examples/l3fwd/l3fwd_fib.c
index 18398492ae..6a21984415 100644
--- a/examples/l3fwd/l3fwd_fib.c
+++ b/examples/l3fwd/l3fwd_fib.c
@@ -253,6 +253,7 @@ fib_main_loop(__rte_u

[PATCH v4 7/8] build: make most device classes optional

[Bruce Richardson]

Apart from ethdev and cryptodev, which have lots of components and tests
which depend on them, we can make the device class libraries optional
without too much work.

This patch marks:
* bbdev,
* compressdev,
* dmadev,
* eventdev,
* mldev,
* rawdev,
* regexdev
optional, and ensures that DPDK - including tests - can be built with
these components disabled.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 lib/meson.build | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/lib/meson.build b/lib/meson.build
index 92cbd6bb66..1a78c8c984 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -67,8 +67,12 @@ libraries = [
 ]
 
 optional_libs = [
+'bbdev',
 'bitratestats',
 'cfgfile',
+'compressdev',
+'dmadev',
+'eventdev',
 'gpudev',
 'graph',
 'gro',
@@ -76,11 +80,14 @@ optional_libs = [
 'jobstats',
 'latencystats',
 'metrics',
+'mldev',
 'node',
 'pdump',
 'pipeline',
 'port',
 'power',
+'rawdev',
+'regexdev',
 'table',
 'vhost',
 ]
-- 
2.39.2

[PATCH v4 8/8] build: expand list of optional libraries

[Bruce Richardson]

With the unit test build now with individual per-file dependencies, we
can more easily expand the list of optional libraries. Add 8 new
libraries to the optional list.

Signed-off-by: Bruce Richardson 
---
 lib/meson.build | 8 
 1 file changed, 8 insertions(+)

diff --git a/lib/meson.build b/lib/meson.build
index 1a78c8c984..0c9dc8b3b6 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -69,25 +69,33 @@ libraries = [
 optional_libs = [
 'bbdev',
 'bitratestats',
+'bpf',
 'cfgfile',
 'compressdev',
+'distributor',
 'dmadev',
+'efd',
 'eventdev',
 'gpudev',
 'graph',
 'gro',
 'gso',
+'ip_frag',
 'jobstats',
 'latencystats',
+'member',
 'metrics',
 'mldev',
 'node',
+'pcapng',
 'pdump',
 'pipeline',
 'port',
 'power',
 'rawdev',
 'regexdev',
+'reorder',
+'sched',
 'table',
 'vhost',
 ]
-- 
2.39.2

Re: [PATCH v4 0/8] expand list of optional libraries

[Bruce Richardson]

On Tue, Aug 15, 2023 at 02:13:12PM +0100, Bruce Richardson wrote:
> This patchset contains changes to DPDK to make the build more configurable
> overall. The changes can be grouped into two areas:
> 
> * Firstly, there are changes to make the build of the unit tests more
>   flexible and maintainable. These 5 patches switch the unit tests from
>   having separate hard-coded lists of files and tests, with lots of
>   conditional logic to add/remove items from the list, to instead being
>   a single list of files and dependencies per-file. The actual lists of
>   test-suites are built up dynamically from the list of files whose
>   dependencies are met and therefore will be part of the build.
>   This rework enables...
> * The second set of patches which expands the list of libraries which
>   can be disabled at build-time, while still giving a successful build
>   of DPDK. Overall, after this set is applied, the list of optional
>   libraries in DPDK is approx doubled in size. [15 additions giving
>   31 optional libs]
> 
> NOTES:
> * as stated in the RFC, the test binary moves in the build as a result
>   of this set - from app/test/dpdk-test to app/dpdk-test.
> * running the suites of tests no longer works after patch 4 until patch
>   5 is applied. If this is an issue, the two patches can be merged on
>   apply. However, the individual tests are still runnable through the
>   test binary - it's only through "meson test" that there are issues, so
>   I'm keeping the patches separate since each is relatively complex in its
>   own right.
> * A number of files in the test directory were not being built, and seem
>   to have never been built with meson. This requires separate work, outside
>   the scope of the patchset to fix.
> * Similarly, some unit tests were never added to test suites. These can
>   now be found by using grep to find the old REGISTER_TEST_COMMAND macro.
>   Again, adding orphan tests to suites is outside the scope of this set.
> 
> V4:
> - improve test suite definition, hopefully improving test pass rate:
>   - drop unnecessary file-prefix for fast tests, since tests are not parallel
>   - always pass driver path flag in "meson test" for shared builds.
> 
Missed applying Morten's ack to the series in V4. Adding it here.

Series-acked-by: Morten Brørup 

[Bug 1274] [20.11 LTS] MPRQ parameters are incorrectly adjusted and not applied properly

[bugzilla]

https://bugs.dpdk.org/show_bug.cgi?id=1274

Bug ID: 1274
   Summary: [20.11 LTS] MPRQ parameters are incorrectly adjusted
and not applied properly
   Product: DPDK
   Version: 20.11
  Hardware: All
OS: All
Status: UNCONFIRMED
  Severity: normal
  Priority: Normal
 Component: ethdev
  Assignee: dev@dpdk.org
  Reporter: alia...@nvidia.com
  Target Milestone: ---

To reproduce:

"""
$ dpdk-testpmd -n 4 -a
:08:00.0,dv_xmeta_en=1,rxqs_min_mprq=4,mprq_en=1,mprq_max_memcpy_len=1606,mprq_log_stride_size=7,mprq_log_stride_num=10,txq_inline_max=472,txq_inline_mpw=401,txqs_min_inline=12,hp_buf_log_sz=17,representor=[0,1]
--iova-mode="va" -- --mbcache=512 -i --nb-cores=8 --rxq=8 --txq=8 --txd=2048
--rxd=2048 --burst=64 --enable-scatter --tx-offloads=0x8000
[..]
Configuring Port 0 (socket 0)
mlx5_pci: Port 0 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (1646)
[..]
Port 0: 10:70:FD:D9:0F:C8
Configuring Port 1 (socket 0)
mlx5_pci: Port 1 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (1646)
[..]
Port 1: A6:68:00:C8:79:FD
Configuring Port 2 (socket 0)
mlx5_pci: Port 2 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (1646)
[..]
Port 2: C6:8F:13:5B:6B:F0
Checking link statuses...
Done
testpmd> port stop all
Stopping ports...
Checking link statuses...
Done
testpmd> port config all max-pkt-len 4096
testpmd> port start all
Configuring Port 0 (socket 0)
mlx5_pci: Port 0 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (4224)
[..]
Port 0: 10:70:FD:D9:0F:C8
Configuring Port 1 (socket 0)
mlx5_pci: Port 1 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (4224)
[..]
Port 1: A6:68:00:C8:79:FD
Configuring Port 2 (socket 0)
mlx5_pci: Port 2 Rx queue 0 size of a stride for Multi-Packet RQ is adjusted to
match the mbuf size (4224)
[..]
mlx5_pci: port 2 failed to allocate a mempool for Multi-Packet RQ, count=1280,
size=526016
mlx5_pci: port 2 Rx queue allocation failed: Cannot allocate memory
Fail to start port 2
Please stop the ports first
"""

OS: Ubuntu 20.04
DPDK: 20.11.9-rc1
Device: ConnectX-6 DX
MLNX_OFED: 23.04-1.1.3.0
Firmware: 22.37.1014

-- 
You are receiving this mail because:
You are the assignee for the bug.

Re: [PATCH v4 4/4] build: enable MSVC specific compiler options

[David Marchand]

Hello Tyler,

On Fri, Aug 11, 2023 at 8:24 PM Tyler Retzlaff
 wrote:
>
> * Enable optional use of C11 atomics support.
> * Enable use of C23 typeof operator.
> * Explicitly force intrinsics when building with MSVC.
> * Disable MSVC C runtime checks.
>
> Signed-off-by: Tyler Retzlaff 
> ---
>  config/meson.build | 8 
>  1 file changed, 8 insertions(+)
>
> diff --git a/config/meson.build b/config/meson.build
> index 821a1c3..839057a 100644
> --- a/config/meson.build
> +++ b/config/meson.build
> @@ -27,6 +27,14 @@ dpdk_conf.set('RTE_EXEC_ENV_' + exec_env.to_upper(), 1)
>  is_ms_compiler = is_windows and (cc.get_id() == 'msvc')
>  is_ms_linker = is_windows and (cc.get_id() == 'clang' or is_ms_compiler)
>
> +# MS compiler (except x86) does not support inline assembly

One last (hopefully) small question, what does this comment apply to?
Forced use of intrinsics?

> +if is_ms_compiler
> +dpdk_conf.set('_CRT_SECURE_NO_WARNINGS', 1)
> +dpdk_conf.set('RTE_FORCE_INTRINSICS', 1)
> +add_project_arguments('/experimental:c11atomics', language: 'c')
> +add_project_arguments('/d1experimental:typeof', language: 'c')
> +endif
> +
>  # set the major version, which might be used by drivers and libraries
>  # depending on the configuration options
>  pver = meson.project_version().split('.')
> --
> 1.8.3.1
>


-- 
David Marchand

RE: 20.11.9 patches review and test

[Ali Alnubani]

> -Original Message-
> From: Luca Boccassi 
> Sent: Monday, August 14, 2023 2:55 PM
> To: Ali Alnubani 
> Cc: sta...@dpdk.org; dev@dpdk.org; John McNamara
> ; Raslan Darawsheh ;
> NBU-Contact-Thomas Monjalon (EXTERNAL) ; Wael
> Abualrub 
> Subject: Re: 20.11.9 patches review and test
> 
> On Mon, 14 Aug 2023 at 12:52, Ali Alnubani  wrote:
> >
> > > -Original Message-
> > > From: luca.bocca...@gmail.com 
> > > Sent: Saturday, July 29, 2023 2:08 AM
> > > To: sta...@dpdk.org
> > > Cc: dev@dpdk.org; Abhishek Marathe
> ;
> > > Ali Alnubani ; benjamin.wal...@intel.com; David
> > > Christensen ; Hemant Agrawal
> > > ; Ian Stokes ; Jerin
> Jacob
> > > ; John McNamara ; Ju-
> > > Hyoung Lee ; Kevin Traynor
> ;
> > > Pei Zhang ; qian.q...@intel.com; Raslan
> Darawsheh
> > > ; NBU-Contact-Thomas Monjalon (EXTERNAL)
> > > ; Yanghang Liu ;
> > > yuan.p...@intel.com; zhaoyan.c...@intel.com
> > > Subject: 20.11.9 patches review and test
> > >
> > > Hi all,
> > >
> > > Here is a list of patches targeted for stable release 20.11.9.
> > >
> > > The planned date for the final release is the 14th of August 2023.
> > >
> > > Please help with testing and validation of your use cases and report
> > > any issues/results with reply-all to this mail. For the final release
> > > the fixes and reported validations will be added to the release notes.
> > >
> > > A release candidate tarball can be found at:
> > >
> > > https://dpdk.org/browse/dpdk-stable/tag/?id=v20.11.9-rc1
> > >
> > > These patches are located at branch 20.11 of dpdk-stable repo:
> > > https://dpdk.org/browse/dpdk-stable/
> > >
> > > Thanks.
> > >
> > > Luca Boccassi
> > >
> > > ---
> >
> >  Hello,
> >
> > Apologies for the last-minute reply.
> >
> > We ran the following functional tests with Nvidia hardware on v20.11.9-rc1:
> > - Basic functionality:
> >   Send and receive multiple types of traffic.
> > - testpmd xstats counter test.
> > - testpmd timestamp test.
> > - Changing/checking link status through testpmd.
> > - rte_flow tests.
> > - Some RSS tests.
> > - VLAN filtering, stripping and insertion tests.
> > - Checksum and TSO tests.
> > - ptype tests.
> > - link_status_interrupt example application tests.
> > - l3fwd-power example application tests.
> > - Multi-process example applications tests.
> > - Hardware LRO tests.
> >
> > Functional tests ran on these cards:
> > - NIC: ConnectX-6 Dx / OS: Ubuntu 20.04 / Driver: MLNX_OFED_LINUX-
> 23.04-1.1.3.0 / Firmware: 22.37.1014
> > - NIC: ConnectX-7 / OS: Ubuntu 20.04 / Driver: MLNX_OFED_LINUX-23.04-
> 1.1.3.0 / Firmware: 28.37.1014
> > - DPU: BlueField-2 / DOCA SW version: 2.0.2 / Firmware: 24.37.1300
> >
> > Build tests (all passed):
> > - Ubuntu 20.04.6 with MLNX_OFED_LINUX-23.04-1.1.3.0.
> > - Ubuntu 20.04.6 with rdma-core master (aba30bd).
> > - Ubuntu 20.04.6 with rdma-core v28.0.
> > - Ubuntu 18.04.6 with rdma-core master (aba30bd) (i386).
> > - Fedora 38 with rdma-core v44.0.
> > - Fedora 39 (Rawhide) with rdma-core v46.0.
> > - OpenSUSE Leap 15.5 with rdma-core v42.0.
> > - Windows Server 2019 with Clang 11.0.0.
> >
> > We see warnings while restarting ports in testpmd application if MPRQ is
> enabled, and we think it's caused by this backported change:
> > net/mlx5: fix MPRQ stride size to accommodate the headroom
> (https://git.dpdk.org/dpdk-stable/commit/?h=81002988e01)
> > This is the only issue we found that might've been caused by changes in this
> release. Hope I can update by tomorrow.
> >
> > We're investigating some internal test failures that were caused by
> environment changes, but we have a high pass rate in our regression testing
> overall.
> >
> > Thanks,
> > Ali
> 
> Thank you, I'll wait for the conclusion on that, if I need to revert
> that change let me know.
> 

Hi Luca,

I created a Bugzilla ticket to track the issues with Multi-Packet Rx queue 
(mprq). We hope to resolve it in 20.11.10:

https://bugs.dpdk.org/show_bug.cgi?id=1274 - [20.11 LTS] Multi-Packet Rx queue 
parameters are incorrectly adjusted and not applied properly

We're ok to release.

Thanks,
Ali

[PATCH v4 0/8] support reinit flow

[okaya]

From: Sinan Kaya 

We want to be able to call rte_eal_init() and rte_eal_cleanup()
APIs back to back for maintanance reasons.

Here is a summary of the code we have seen so far:

1. some code support getting called multiple times by keeping
a static variable.
2. some code initializes once but never clean up after them and
don't have a cleanup API.
3. some code assumes that they only get called once during the
lifecycle of the process.

Most changes in this patch center around following the #1 design
principle.

Why?

It is not always ideal to reinitialize a DPDK process. Memory needs
to be reinitialized, hugetables need to warm up etc.

Changed from

v1:
Fix checkpatch warnings

v2:
rebase to most recent DPDK.

v3:
pick up Stephen's "eal: cleanup plugins data" as a pre-requisite
patch.

Graham Whyte (1):
  eal: fixes for re-initialization issues

Sinan Kaya (6):
  tailq: skip init if already initialized
  eal_memzone: bail out on initialized
  memseg: init once
  eal_memory: skip initialization
  eal_interrupts: don't reinitialize threads
  eal: initialize worker threads once

Stephen Hemminger (1):
  eal: cleanup plugins data

 lib/eal/common/eal_common_memory.c  |  5 +++
 lib/eal/common/eal_common_memzone.c |  7 +++
 lib/eal/common/eal_common_options.c | 20 +
 lib/eal/common/eal_common_tailqs.c  | 20 ++---
 lib/eal/common/eal_options.h|  1 +
 lib/eal/common/malloc_heap.c|  7 +++
 lib/eal/linux/eal.c | 66 -
 lib/eal/linux/eal_interrupts.c  |  7 +++
 lib/eal/linux/eal_memory.c  | 12 +-
 9 files changed, 108 insertions(+), 37 deletions(-)

-- 
2.25.1

[PATCH v4 1/8] eal: cleanup plugins data

[okaya]

From: Stephen Hemminger 

When rte_eal_cleanup is called walk through the list of shared
objects loaded, and close them and free the data structure.

Signed-off-by: Stephen Hemminger 
Signed-off-by: Sinan Kaya 
---
 lib/eal/common/eal_common_options.c | 12 
 lib/eal/common/eal_options.h|  1 +
 lib/eal/linux/eal.c |  2 ++
 3 files changed, 15 insertions(+)

diff --git a/lib/eal/common/eal_common_options.c 
b/lib/eal/common/eal_common_options.c
index 062f1d8d9c..209b6edd76 100644
--- a/lib/eal/common/eal_common_options.c
+++ b/lib/eal/common/eal_common_options.c
@@ -244,6 +244,18 @@ eal_save_args(int argc, char **argv)
 }
 #endif
 
+void
+eal_plugins_cleanup(void)
+{
+   struct shared_driver *solib, *tmp;
+
+   RTE_TAILQ_FOREACH_SAFE(solib, &solib_list, next, tmp) {
+   if (solib->lib_handle)
+   dlclose(solib->lib_handle);
+   free(solib);
+   }
+}
+
 static int
 eal_option_device_add(enum rte_devtype type, const char *optarg)
 {
diff --git a/lib/eal/common/eal_options.h b/lib/eal/common/eal_options.h
index 3cc9cb6412..ddbaafc4f1 100644
--- a/lib/eal/common/eal_options.h
+++ b/lib/eal/common/eal_options.h
@@ -105,6 +105,7 @@ int eal_check_common_options(struct internal_config 
*internal_cfg);
 void eal_common_usage(void);
 enum rte_proc_type_t eal_proc_type_detect(void);
 int eal_plugins_init(void);
+void eal_plugins_cleanup(void);
 int eal_save_args(int argc, char **argv);
 int handle_eal_info_request(const char *cmd, const char *params __rte_unused,
struct rte_tel_data *d);
diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index c6efd92014..dee649bab3 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -1398,6 +1398,8 @@ rte_eal_cleanup(void)
eal_trace_fini();
eal_mp_dev_hotplug_cleanup();
rte_eal_alarm_cleanup();
+   eal_plugins_cleanup();
+
/* after this point, any DPDK pointers will become dangling */
rte_eal_memory_detach();
rte_eal_malloc_heap_cleanup();
-- 
2.25.1

[PATCH v4 2/8] eal: fixes for re-initialization issues

[okaya]

From: Graham Whyte 

reinitialize the solib link list and clean the globals holding
state for parsing.

Signed-off-by: Sinan Kaya 
Signed-off-by: Graham Whyte 
---
 lib/eal/common/eal_common_options.c | 8 
 lib/eal/linux/eal.c | 4 +++-
 2 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/lib/eal/common/eal_common_options.c 
b/lib/eal/common/eal_common_options.c
index 209b6edd76..6042de009d 100644
--- a/lib/eal/common/eal_common_options.c
+++ b/lib/eal/common/eal_common_options.c
@@ -127,6 +127,7 @@ TAILQ_HEAD_INITIALIZER(solib_list);
 static const char *default_solib_dir = RTE_EAL_PMD_PATH;
 #endif
 
+
 /*
  * Stringified version of solib path used by dpdk-pmdinfo.py
  * Note: PLEASE DO NOT ALTER THIS without making a corresponding
@@ -254,6 +255,13 @@ eal_plugins_cleanup(void)
dlclose(solib->lib_handle);
free(solib);
}
+
+   /* Reinitialize solib_list */
+   TAILQ_INIT(&solib_list);
+
+   main_lcore_parsed = 0;
+   mem_parsed = 0;
+   core_parsed = 0;
 }
 
 static int
diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index dee649bab3..584c78e640 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -79,6 +79,8 @@ struct lcore_config lcore_config[RTE_MAX_LCORE];
 int rte_cycles_vmware_tsc_map;
 
 
+static uint32_t run_once;
+
 int
 eal_clean_runtime_dir(void)
 {
@@ -505,6 +507,7 @@ eal_parse_socket_arg(char *strval, volatile uint64_t 
*socket_arg)
socket_arg[i] = val;
}
 
+   __atomic_store_n(&run_once, 0, __ATOMIC_RELAXED);
return 0;
 }
 
@@ -967,7 +970,6 @@ int
 rte_eal_init(int argc, char **argv)
 {
int i, fctret, ret;
-   static uint32_t run_once;
uint32_t has_run = 0;
char cpuset[RTE_CPU_AFFINITY_STR_LEN];
char thread_name[RTE_MAX_THREAD_NAME_LEN];
-- 
2.25.1

[PATCH v4 3/8] tailq: skip init if already initialized

[okaya]

From: Sinan Kaya 

Allows tailq to be reinitialied multiple times
by looking up previously registered tailqs

Signed-off-by: Sinan Kaya 
---
 lib/eal/common/eal_common_tailqs.c | 20 +---
 1 file changed, 13 insertions(+), 7 deletions(-)

diff --git a/lib/eal/common/eal_common_tailqs.c 
b/lib/eal/common/eal_common_tailqs.c
index 580fbf24bc..75c0235438 100644
--- a/lib/eal/common/eal_common_tailqs.c
+++ b/lib/eal/common/eal_common_tailqs.c
@@ -73,9 +73,10 @@ rte_eal_tailq_create(const char *name)
strlcpy(head->name, name, sizeof(head->name) - 1);
TAILQ_INIT(&head->tailq_head);
rte_tailqs_count++;
+   return head;
}
 
-   return head;
+   return rte_eal_tailq_lookup(name);
 }
 
 /* local register, used to store "early" tailqs before rte_eal_init() and to
@@ -99,7 +100,9 @@ rte_eal_tailq_update(struct rte_tailq_elem *t)
 {
if (rte_eal_process_type() == RTE_PROC_PRIMARY) {
/* primary process is the only one that creates */
-   t->head = rte_eal_tailq_create(t->name);
+   t->head = rte_eal_tailq_lookup(t->name);
+   if (t->head == NULL)
+   t->head = rte_eal_tailq_create(t->name);
} else {
t->head = rte_eal_tailq_lookup(t->name);
}
@@ -108,15 +111,13 @@ rte_eal_tailq_update(struct rte_tailq_elem *t)
 int
 rte_eal_tailq_register(struct rte_tailq_elem *t)
 {
-   if (rte_eal_tailq_local_register(t) < 0) {
-   RTE_LOG(ERR, EAL,
-   "%s tailq is already registered\n", t->name);
-   goto error;
-   }
+   rte_eal_tailq_local_register(t);
 
/* if a register happens after rte_eal_tailqs_init(), then we can update
 * tailq head */
if (rte_tailqs_count >= 0) {
+   RTE_LOG(INFO, EAL,
+   "%s tailq is registered\n", t->name);
rte_eal_tailq_update(t);
if (t->head == NULL) {
RTE_LOG(ERR, EAL,
@@ -138,6 +139,11 @@ rte_eal_tailqs_init(void)
 {
struct rte_tailq_elem *t;
 
+   if (rte_tailqs_count > 0) {
+   RTE_LOG(INFO, EAL, "tailq already initialized\n");
+   return 0;
+   }
+
rte_tailqs_count = 0;
 
TAILQ_FOREACH(t, &rte_tailq_elem_head, next) {
-- 
2.25.1

[PATCH v4 4/8] eal_memzone: bail out on initialized

[okaya]

From: Sinan Kaya 

Initialize memzone once and bail out if someone calls init
multiple times.

Signed-off-by: Sinan Kaya 
---
 lib/eal/common/eal_common_memzone.c | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/lib/eal/common/eal_common_memzone.c 
b/lib/eal/common/eal_common_memzone.c
index 1f3e701499..6645ccfe83 100644
--- a/lib/eal/common/eal_common_memzone.c
+++ b/lib/eal/common/eal_common_memzone.c
@@ -22,6 +22,8 @@
 #include "eal_private.h"
 #include "eal_memcfg.h"
 
+static bool memzone_initialized;
+
 /* Default count used until rte_memzone_max_set() is called */
 #define DEFAULT_MAX_MEMZONE_COUNT 2560
 
@@ -426,6 +428,9 @@ rte_eal_memzone_init(void)
struct rte_mem_config *mcfg;
int ret = 0;
 
+   if (memzone_initialized)
+   return 0;
+
/* get pointer to global configuration */
mcfg = rte_eal_get_configuration()->mem_config;
 
@@ -444,6 +449,8 @@ rte_eal_memzone_init(void)
 
rte_rwlock_write_unlock(&mcfg->mlock);
 
+   memzone_initialized = true;
+
return ret;
 }
 
-- 
2.25.1

[PATCH v4 5/8] memseg: init once

[okaya]

From: Sinan Kaya 

Initialize memory segments just once and bail out if called
multiple times.

Signed-off-by: Sinan Kaya 
---
 lib/eal/linux/eal_memory.c | 12 +++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/lib/eal/linux/eal_memory.c b/lib/eal/linux/eal_memory.c
index 9b6f08fba8..df0aa9ccc7 100644
--- a/lib/eal/linux/eal_memory.c
+++ b/lib/eal/linux/eal_memory.c
@@ -57,6 +57,7 @@
  */
 
 static int phys_addrs_available = -1;
+static bool memseg_initialized;
 
 #define RANDOMIZE_VA_SPACE_FILE "/proc/sys/kernel/randomize_va_space"
 
@@ -1920,6 +1921,10 @@ rte_eal_memseg_init(void)
 {
/* increase rlimit to maximum */
struct rlimit lim;
+   int ret;
+
+   if (memseg_initialized)
+   return 0;
 
 #ifndef RTE_EAL_NUMA_AWARE_HUGEPAGES
const struct internal_config *internal_conf =
@@ -1948,11 +1953,16 @@ rte_eal_memseg_init(void)
}
 #endif
 
-   return rte_eal_process_type() == RTE_PROC_PRIMARY ?
+   ret = rte_eal_process_type() == RTE_PROC_PRIMARY ?
 #ifndef RTE_ARCH_64
memseg_primary_init_32() :
 #else
memseg_primary_init() :
 #endif
memseg_secondary_init();
+
+   if (!ret)
+   memseg_initialized = true;
+
+   return ret;
 }
-- 
2.25.1

[PATCH v4 6/8] eal_memory: skip initialization

[okaya]

From: Sinan Kaya 

Initialize heap area just once.

Signed-off-by: Sinan Kaya 
---
 lib/eal/common/eal_common_memory.c | 5 +
 lib/eal/common/malloc_heap.c   | 7 +++
 2 files changed, 12 insertions(+)

diff --git a/lib/eal/common/eal_common_memory.c 
b/lib/eal/common/eal_common_memory.c
index d9433db623..4c68de1054 100644
--- a/lib/eal/common/eal_common_memory.c
+++ b/lib/eal/common/eal_common_memory.c
@@ -41,6 +41,7 @@
 
 static void *next_baseaddr;
 static uint64_t system_page_sz;
+static bool memory_initialized;
 
 #define MAX_MMAP_WITH_DEFINED_ADDR_TRIES 5
 void *
@@ -1084,6 +1085,9 @@ rte_eal_memory_init(void)
eal_get_internal_configuration();
int retval;
 
+   if (memory_initialized)
+   return 0;
+
RTE_LOG(DEBUG, EAL, "Setting up physically contiguous memory...\n");
 
if (rte_eal_memseg_init() < 0)
@@ -1101,6 +1105,7 @@ rte_eal_memory_init(void)
if (internal_conf->no_shconf == 0 && rte_eal_memdevice_init() < 0)
goto fail;
 
+   memory_initialized = true;
return 0;
 fail:
return -1;
diff --git a/lib/eal/common/malloc_heap.c b/lib/eal/common/malloc_heap.c
index 6b6cf9174c..8d05a2ab2e 100644
--- a/lib/eal/common/malloc_heap.c
+++ b/lib/eal/common/malloc_heap.c
@@ -31,6 +31,8 @@
 #define CONST_MAX(a, b) (a > b ? a : b) /* RTE_MAX is not a constant */
 #define EXTERNAL_HEAP_MIN_SOCKET_ID (CONST_MAX((1 << 8), RTE_MAX_NUMA_NODES))
 
+static bool heap_initialized;
+
 static unsigned
 check_hugepage_sz(unsigned flags, uint64_t hugepage_sz)
 {
@@ -1410,6 +1412,9 @@ rte_eal_malloc_heap_init(void)
const struct internal_config *internal_conf =
eal_get_internal_configuration();
 
+   if (heap_initialized)
+   return 0;
+
if (internal_conf->match_allocations)
RTE_LOG(DEBUG, EAL, "Hugepages will be freed exactly as 
allocated.\n");
 
@@ -1449,6 +1454,8 @@ int rte_eal_malloc_heap_populate(void)
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
 
+   heap_initialized = true;
+
/* add all IOVA-contiguous areas to the heap */
return rte_memseg_contig_walk(malloc_add_seg, NULL);
 }
-- 
2.25.1

[PATCH v4 7/8] eal_interrupts: don't reinitialize threads

[okaya]

From: Sinan Kaya 

Initialize interrupt thread once and keep a flag
for further init.

Signed-off-by: Sinan Kaya 
---
 lib/eal/linux/eal_interrupts.c | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/lib/eal/linux/eal_interrupts.c b/lib/eal/linux/eal_interrupts.c
index c9881143be..6a35b4aebd 100644
--- a/lib/eal/linux/eal_interrupts.c
+++ b/lib/eal/linux/eal_interrupts.c
@@ -91,6 +91,9 @@ static struct rte_intr_source_list intr_sources;
 /* interrupt handling thread */
 static pthread_t intr_thread;
 
+/* flag for initialization */
+static bool intr_initialized;
+
 /* VFIO interrupts */
 #ifdef VFIO_PRESENT
 
@@ -1175,6 +1178,9 @@ rte_eal_intr_init(void)
 {
int ret = 0;
 
+   if (intr_initialized)
+   return 0;
+
/* init the global interrupt source head */
TAILQ_INIT(&intr_sources);
 
@@ -1196,6 +1202,7 @@ rte_eal_intr_init(void)
"Failed to create thread for interrupt handling\n");
}
 
+   intr_initialized = true;
return ret;
 }
 
-- 
2.25.1

[PATCH v4 8/8] eal: initialize worker threads once

[okaya]

From: Sinan Kaya 

Initialize worker threads once and keep a flag
for other init calls.

Signed-off-by: Sinan Kaya 
---
 lib/eal/linux/eal.c | 60 -
 1 file changed, 32 insertions(+), 28 deletions(-)

diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index 584c78e640..1305e1df54 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -80,6 +80,7 @@ int rte_cycles_vmware_tsc_map;
 
 
 static uint32_t run_once;
+static bool worker_initialized;
 
 int
 eal_clean_runtime_dir(void)
@@ -1254,42 +1255,45 @@ rte_eal_init(int argc, char **argv)
config->main_lcore, (uintptr_t)pthread_self(), cpuset,
ret == 0 ? "" : "...");
 
-   RTE_LCORE_FOREACH_WORKER(i) {
+   if (worker_initialized == false) {
+   RTE_LCORE_FOREACH_WORKER(i) {
 
/*
 * create communication pipes between main thread
 * and children
 */
-   if (pipe(lcore_config[i].pipe_main2worker) < 0)
-   rte_panic("Cannot create pipe\n");
-   if (pipe(lcore_config[i].pipe_worker2main) < 0)
-   rte_panic("Cannot create pipe\n");
-
-   lcore_config[i].state = WAIT;
-
-   /* create a thread for each lcore */
-   ret = eal_worker_thread_create(i);
-   if (ret != 0)
-   rte_panic("Cannot create thread\n");
-
-   /* Set thread_name for aid in debugging. */
-   snprintf(thread_name, sizeof(thread_name),
-   "rte-worker-%d", i);
-   rte_thread_set_name(lcore_config[i].thread_id, thread_name);
+   if (pipe(lcore_config[i].pipe_main2worker) < 0)
+   rte_panic("Cannot create pipe\n");
+   if (pipe(lcore_config[i].pipe_worker2main) < 0)
+   rte_panic("Cannot create pipe\n");
+
+   lcore_config[i].state = WAIT;
+
+   /* create a thread for each lcore */
+   ret = eal_worker_thread_create(i);
+   if (ret != 0)
+   rte_panic("Cannot create thread\n");
+
+   /* Set thread_name for aid in debugging. */
+   snprintf(thread_name, sizeof(thread_name),
+   "rte-worker-%d", i);
+   rte_thread_set_name(lcore_config[i].thread_id, 
thread_name);
+
+   ret = 
rte_thread_set_affinity_by_id(lcore_config[i].thread_id,
+   &lcore_config[i].cpuset);
+   if (ret != 0)
+   rte_panic("Cannot set affinity\n");
+   }
 
-   ret = rte_thread_set_affinity_by_id(lcore_config[i].thread_id,
-   &lcore_config[i].cpuset);
-   if (ret != 0)
-   rte_panic("Cannot set affinity\n");
+   /*
+* Launch a dummy function on all worker lcores, so that main 
lcore
+* knows they are all ready when this function returns.
+*/
+   rte_eal_mp_remote_launch(sync_func, NULL, SKIP_MAIN);
+   rte_eal_mp_wait_lcore();
+   worker_initialized = true;
}
 
-   /*
-* Launch a dummy function on all worker lcores, so that main lcore
-* knows they are all ready when this function returns.
-*/
-   rte_eal_mp_remote_launch(sync_func, NULL, SKIP_MAIN);
-   rte_eal_mp_wait_lcore();
-
/* initialize services so vdevs register service during bus_probe. */
ret = rte_service_init();
if (ret) {
-- 
2.25.1

Re: 20.11.9 patches review and test

[Luca Boccassi]

On Tue, 15 Aug 2023 at 14:35, Ali Alnubani  wrote:
>
> > -Original Message-
> > From: Luca Boccassi 
> > Sent: Monday, August 14, 2023 2:55 PM
> > To: Ali Alnubani 
> > Cc: sta...@dpdk.org; dev@dpdk.org; John McNamara
> > ; Raslan Darawsheh ;
> > NBU-Contact-Thomas Monjalon (EXTERNAL) ; Wael
> > Abualrub 
> > Subject: Re: 20.11.9 patches review and test
> >
> > On Mon, 14 Aug 2023 at 12:52, Ali Alnubani  wrote:
> > >
> > > > -Original Message-
> > > > From: luca.bocca...@gmail.com 
> > > > Sent: Saturday, July 29, 2023 2:08 AM
> > > > To: sta...@dpdk.org
> > > > Cc: dev@dpdk.org; Abhishek Marathe
> > ;
> > > > Ali Alnubani ; benjamin.wal...@intel.com; David
> > > > Christensen ; Hemant Agrawal
> > > > ; Ian Stokes ; Jerin
> > Jacob
> > > > ; John McNamara ; Ju-
> > > > Hyoung Lee ; Kevin Traynor
> > ;
> > > > Pei Zhang ; qian.q...@intel.com; Raslan
> > Darawsheh
> > > > ; NBU-Contact-Thomas Monjalon (EXTERNAL)
> > > > ; Yanghang Liu ;
> > > > yuan.p...@intel.com; zhaoyan.c...@intel.com
> > > > Subject: 20.11.9 patches review and test
> > > >
> > > > Hi all,
> > > >
> > > > Here is a list of patches targeted for stable release 20.11.9.
> > > >
> > > > The planned date for the final release is the 14th of August 2023.
> > > >
> > > > Please help with testing and validation of your use cases and report
> > > > any issues/results with reply-all to this mail. For the final release
> > > > the fixes and reported validations will be added to the release notes.
> > > >
> > > > A release candidate tarball can be found at:
> > > >
> > > > https://dpdk.org/browse/dpdk-stable/tag/?id=v20.11.9-rc1
> > > >
> > > > These patches are located at branch 20.11 of dpdk-stable repo:
> > > > https://dpdk.org/browse/dpdk-stable/
> > > >
> > > > Thanks.
> > > >
> > > > Luca Boccassi
> > > >
> > > > ---
> > >
> > >  Hello,
> > >
> > > Apologies for the last-minute reply.
> > >
> > > We ran the following functional tests with Nvidia hardware on 
> > > v20.11.9-rc1:
> > > - Basic functionality:
> > >   Send and receive multiple types of traffic.
> > > - testpmd xstats counter test.
> > > - testpmd timestamp test.
> > > - Changing/checking link status through testpmd.
> > > - rte_flow tests.
> > > - Some RSS tests.
> > > - VLAN filtering, stripping and insertion tests.
> > > - Checksum and TSO tests.
> > > - ptype tests.
> > > - link_status_interrupt example application tests.
> > > - l3fwd-power example application tests.
> > > - Multi-process example applications tests.
> > > - Hardware LRO tests.
> > >
> > > Functional tests ran on these cards:
> > > - NIC: ConnectX-6 Dx / OS: Ubuntu 20.04 / Driver: MLNX_OFED_LINUX-
> > 23.04-1.1.3.0 / Firmware: 22.37.1014
> > > - NIC: ConnectX-7 / OS: Ubuntu 20.04 / Driver: MLNX_OFED_LINUX-23.04-
> > 1.1.3.0 / Firmware: 28.37.1014
> > > - DPU: BlueField-2 / DOCA SW version: 2.0.2 / Firmware: 24.37.1300
> > >
> > > Build tests (all passed):
> > > - Ubuntu 20.04.6 with MLNX_OFED_LINUX-23.04-1.1.3.0.
> > > - Ubuntu 20.04.6 with rdma-core master (aba30bd).
> > > - Ubuntu 20.04.6 with rdma-core v28.0.
> > > - Ubuntu 18.04.6 with rdma-core master (aba30bd) (i386).
> > > - Fedora 38 with rdma-core v44.0.
> > > - Fedora 39 (Rawhide) with rdma-core v46.0.
> > > - OpenSUSE Leap 15.5 with rdma-core v42.0.
> > > - Windows Server 2019 with Clang 11.0.0.
> > >
> > > We see warnings while restarting ports in testpmd application if MPRQ is
> > enabled, and we think it's caused by this backported change:
> > > net/mlx5: fix MPRQ stride size to accommodate the headroom
> > (https://git.dpdk.org/dpdk-stable/commit/?h=81002988e01)
> > > This is the only issue we found that might've been caused by changes in 
> > > this
> > release. Hope I can update by tomorrow.
> > >
> > > We're investigating some internal test failures that were caused by
> > environment changes, but we have a high pass rate in our regression testing
> > overall.
> > >
> > > Thanks,
> > > Ali
> >
> > Thank you, I'll wait for the conclusion on that, if I need to revert
> > that change let me know.
> >
>
> Hi Luca,
>
> I created a Bugzilla ticket to track the issues with Multi-Packet Rx queue 
> (mprq). We hope to resolve it in 20.11.10:
>
> https://bugs.dpdk.org/show_bug.cgi?id=1274 - [20.11 LTS] Multi-Packet Rx 
> queue parameters are incorrectly adjusted and not applied properly
>
> We're ok to release.
>
> Thanks,
> Ali

Sounds good, thank you.

Kind regards,
Luca Boccassi

[PATCH] app/dma-perf: fix physical address seg-fault

[Vipin Varghese]

do_cpu_mem_copy uses DPDK API rte_mbuf_data_iova to return
the start of the virtual address for both src and dst.
But in case of iova mode set as PA, this results in seg-fault.
This is because rte_memcpy VA address and not PA.

this fix checks the mode and invoked rte_memcpy accrodingly.

Bugzilla ID: 1269
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
Cc: cheng1.ji...@intel.com

Cc: sta...@dpdk.org

Signed-off-by: Vipin Varghese 
---

tested for both va and pa

CMD:
PA: dpdk-test-dma-perf --iova-mode=pa  -- --config test.ini
VA: dpdk-test-dma-perf --iova-mode=va  -- --config test.ini
DC: dpdk-test-dma-perf --iova-mode=dc  -- --config test.ini

Log: fails for dc mode `EAL: invalid parameters for --iova-mode`

test.ini:
```
[case1]
type=CPU_MEM_COPY
mem_size=10
buf_size=64,8192,2,MUL
src_numa_node=0
dst_numa_node=0
cache_flush=0
test_seconds=2
lcore = 7
eal_args=--in-memory --no-pci
```
---
 app/test-dma-perf/benchmark.c | 36 ++-
 1 file changed, 27 insertions(+), 9 deletions(-)

diff --git a/app/test-dma-perf/benchmark.c b/app/test-dma-perf/benchmark.c
index 0601e0d171..73200e1935 100644
--- a/app/test-dma-perf/benchmark.c
+++ b/app/test-dma-perf/benchmark.c
@@ -279,6 +279,10 @@ do_cpu_mem_copy(void *p)
struct rte_mbuf **srcs = para->srcs;
struct rte_mbuf **dsts = para->dsts;
uint32_t i;
+   bool isAddrPaMode = false;
+
+   if (rte_eal_iova_mode() == RTE_IOVA_PA)
+   isAddrPaMode = true;
 
worker_info->stop_flag = false;
worker_info->ready_flag = true;
@@ -286,16 +290,30 @@ do_cpu_mem_copy(void *p)
while (!worker_info->start_flag)
;
 
-   while (1) {
-   for (i = 0; i < nr_buf; i++) {
-   /* copy buffer form src to dst */
-   rte_memcpy((void 
*)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
-   (void *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
-   (size_t)buf_size);
-   worker_info->total_cpl++;
+   if (true == isAddrPaMode) {
+   while (1) {
+   for (i = 0; i < nr_buf; i++) {
+   /* copy buffer form src to dst */
+   rte_memcpy((void *)dsts[i],
+   (void *)srcs[i],
+   (size_t)buf_size);
+   worker_info->total_cpl++;
+   }
+   if (worker_info->stop_flag)
+   break;
+   }
+   } else {
+   while (1) {
+   for (i = 0; i < nr_buf; i++) {
+   /* copy buffer form src to dst */
+   rte_memcpy((void 
*)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
+   (void 
*)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
+   (size_t)buf_size);
+   worker_info->total_cpl++;
+   }
+   if (worker_info->stop_flag)
+   break;
}
-   if (worker_info->stop_flag)
-   break;
}
 
return 0;
-- 
2.34.1

[PATCH] app/dma-perf: fix physical address seg-fault

[Vipin Varghese]

do_cpu_mem_copy uses DPDK API rte_mbuf_data_iova to return
the start of the virtual address for both src and dst.
But in case of iova mode set as PA, this results in seg-fault.
This is because rte_memcpy VA address and not PA.

this fix checks the iova mode and invokes rte_memcpy with the right
arguments.

Bugzilla ID: 1269
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
Cc: cheng1.ji...@intel.com

Cc: sta...@dpdk.org

Signed-off-by: Vipin Varghese 
---

tested for both va and pa

CMD:
PA: dpdk-test-dma-perf --iova-mode=pa  -- --config test.ini
VA: dpdk-test-dma-perf --iova-mode=va  -- --config test.ini
DC: dpdk-test-dma-perf --iova-mode=dc  -- --config test.ini

Log: fails for dc mode `EAL: invalid parameters for --iova-mode`

test.ini:
```
[case1]
type=CPU_MEM_COPY
mem_size=10
buf_size=64,8192,2,MUL
src_numa_node=0
dst_numa_node=0
cache_flush=0
test_seconds=2
lcore = 7
eal_args=--in-memory --no-pci
```
---
 app/test-dma-perf/benchmark.c | 36 ++-
 1 file changed, 27 insertions(+), 9 deletions(-)

diff --git a/app/test-dma-perf/benchmark.c b/app/test-dma-perf/benchmark.c
index 0601e0d171..73200e1935 100644
--- a/app/test-dma-perf/benchmark.c
+++ b/app/test-dma-perf/benchmark.c
@@ -279,6 +279,10 @@ do_cpu_mem_copy(void *p)
struct rte_mbuf **srcs = para->srcs;
struct rte_mbuf **dsts = para->dsts;
uint32_t i;
+   bool isAddrPaMode = false;
+
+   if (rte_eal_iova_mode() == RTE_IOVA_PA)
+   isAddrPaMode = true;
 
worker_info->stop_flag = false;
worker_info->ready_flag = true;
@@ -286,16 +290,30 @@ do_cpu_mem_copy(void *p)
while (!worker_info->start_flag)
;
 
-   while (1) {
-   for (i = 0; i < nr_buf; i++) {
-   /* copy buffer form src to dst */
-   rte_memcpy((void 
*)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
-   (void *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
-   (size_t)buf_size);
-   worker_info->total_cpl++;
+   if (true == isAddrPaMode) {
+   while (1) {
+   for (i = 0; i < nr_buf; i++) {
+   /* copy buffer form src to dst */
+   rte_memcpy((void *)dsts[i],
+   (void *)srcs[i],
+   (size_t)buf_size);
+   worker_info->total_cpl++;
+   }
+   if (worker_info->stop_flag)
+   break;
+   }
+   } else {
+   while (1) {
+   for (i = 0; i < nr_buf; i++) {
+   /* copy buffer form src to dst */
+   rte_memcpy((void 
*)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
+   (void 
*)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
+   (size_t)buf_size);
+   worker_info->total_cpl++;
+   }
+   if (worker_info->stop_flag)
+   break;
}
-   if (worker_info->stop_flag)
-   break;
}
 
return 0;
-- 
2.34.1

[PATCH v5 10/10] build: expand list of optional libraries

[Bruce Richardson]

With the unit test build now with individual per-file dependencies, we
can more easily expand the list of optional libraries. Add 8 new
libraries to the optional list.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 lib/meson.build | 8 
 1 file changed, 8 insertions(+)

diff --git a/lib/meson.build b/lib/meson.build
index 1a78c8c984..0c9dc8b3b6 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -69,25 +69,33 @@ libraries = [
 optional_libs = [
 'bbdev',
 'bitratestats',
+'bpf',
 'cfgfile',
 'compressdev',
+'distributor',
 'dmadev',
+'efd',
 'eventdev',
 'gpudev',
 'graph',
 'gro',
 'gso',
+'ip_frag',
 'jobstats',
 'latencystats',
+'member',
 'metrics',
 'mldev',
 'node',
+'pcapng',
 'pdump',
 'pipeline',
 'port',
 'power',
 'rawdev',
 'regexdev',
+'reorder',
+'sched',
 'table',
 'vhost',
 ]
-- 
2.39.2

RE: [PATCH] app/dma-perf: fix physical address seg-fault

[Varghese, Vipin]

[AMD Official Use Only - General]

Apologies, marking this as `NA`. After recheck of this logic without use of ` 
rte_mbuf_data_iova` will result in mbuf meta-data corruption.

Need to fix this in a different way.

> -Original Message-
> From: Vipin Varghese 
> Sent: Tuesday, August 15, 2023 8:41 PM
> To: tho...@monjalon.net; dev@dpdk.org
> Cc: Yigit, Ferruh ; cheng1.ji...@intel.com;
> sta...@dpdk.org
> Subject: [PATCH] app/dma-perf: fix physical address seg-fault
>
> Caution: This message originated from an External Source. Use proper caution
> when opening attachments, clicking links, or responding.
>
>
> do_cpu_mem_copy uses DPDK API rte_mbuf_data_iova to return the start of
> the virtual address for both src and dst.
> But in case of iova mode set as PA, this results in seg-fault.
> This is because rte_memcpy VA address and not PA.
>
> this fix checks the iova mode and invokes rte_memcpy with the right
> arguments.
>
> Bugzilla ID: 1269
> Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
> Cc: cheng1.ji...@intel.com
>
> Cc: sta...@dpdk.org
>
> Signed-off-by: Vipin Varghese 
> ---
>
> tested for both va and pa
>
> CMD:
> PA: dpdk-test-dma-perf --iova-mode=pa  -- --config test.ini
> VA: dpdk-test-dma-perf --iova-mode=va  -- --config test.ini
> DC: dpdk-test-dma-perf --iova-mode=dc  -- --config test.ini
>
> Log: fails for dc mode `EAL: invalid parameters for --iova-mode`
>
> test.ini:
> ```
> [case1]
> type=CPU_MEM_COPY
> mem_size=10
> buf_size=64,8192,2,MUL
> src_numa_node=0
> dst_numa_node=0
> cache_flush=0
> test_seconds=2
> lcore = 7
> eal_args=--in-memory --no-pci
> ```
> ---
>  app/test-dma-perf/benchmark.c | 36 ++-
>  1 file changed, 27 insertions(+), 9 deletions(-)
>
> diff --git a/app/test-dma-perf/benchmark.c b/app/test-dma-perf/benchmark.c
> index 0601e0d171..73200e1935 100644
> --- a/app/test-dma-perf/benchmark.c
> +++ b/app/test-dma-perf/benchmark.c
> @@ -279,6 +279,10 @@ do_cpu_mem_copy(void *p)
> struct rte_mbuf **srcs = para->srcs;
> struct rte_mbuf **dsts = para->dsts;
> uint32_t i;
> +   bool isAddrPaMode = false;
> +
> +   if (rte_eal_iova_mode() == RTE_IOVA_PA)
> +   isAddrPaMode = true;
>
> worker_info->stop_flag = false;
> worker_info->ready_flag = true;
> @@ -286,16 +290,30 @@ do_cpu_mem_copy(void *p)
> while (!worker_info->start_flag)
> ;
>
> -   while (1) {
> -   for (i = 0; i < nr_buf; i++) {
> -   /* copy buffer form src to dst */
> -   rte_memcpy((void 
> *)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
> -   (void 
> *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
> -   (size_t)buf_size);
> -   worker_info->total_cpl++;
> +   if (true == isAddrPaMode) {
> +   while (1) {
> +   for (i = 0; i < nr_buf; i++) {
> +   /* copy buffer form src to dst */
> +   rte_memcpy((void *)dsts[i],
> +   (void *)srcs[i],
> +   (size_t)buf_size);
> +   worker_info->total_cpl++;
> +   }
> +   if (worker_info->stop_flag)
> +   break;
> +   }
> +   } else {
> +   while (1) {
> +   for (i = 0; i < nr_buf; i++) {
> +   /* copy buffer form src to dst */
> +   rte_memcpy((void
> *)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
> +   (void 
> *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
> +   (size_t)buf_size);
> +   worker_info->total_cpl++;
> +   }
> +   if (worker_info->stop_flag)
> +   break;
> }
> -   if (worker_info->stop_flag)
> -   break;
> }
>
> return 0;
> --
> 2.34.1

[PATCH v5 01/10] app/test: add new macros for various test types

[Bruce Richardson]

Rather than just registering all tests using a single generic macro,
add macros which identify the test as being of a particular type.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 app/test/test.h | 9 -
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/app/test/test.h b/app/test/test.h
index 85f57efbc6..a91ded76af 100644
--- a/app/test/test.h
+++ b/app/test/test.h
@@ -190,7 +190,7 @@ struct test_command {
 
 void add_test_command(struct test_command *t);
 
-/* Register a test function with its command string */
+/* Register a test function with its command string. Should not be used 
directly */
 #define REGISTER_TEST_COMMAND(cmd, func) \
static struct test_command test_struct_##cmd = { \
.command = RTE_STR(cmd), \
@@ -201,4 +201,11 @@ void add_test_command(struct test_command *t);
add_test_command(&test_struct_##cmd); \
}
 
+/* Register a test function as a particular type.
+ * These can be used to build up test suites automatically
+ */
+#define REGISTER_FAST_TEST(cmd, no_huge, ASan, func)  
REGISTER_TEST_COMMAND(cmd, func)
+#define REGISTER_PERF_TEST REGISTER_TEST_COMMAND
+#define REGISTER_DRIVER_TEST REGISTER_TEST_COMMAND
+
 #endif
-- 
2.39.2

[PATCH v5 00/10] expand list of optional libraries

[Bruce Richardson]

This patchset contains changes to DPDK to make the build more configurable
overall. The changes can be grouped into two areas:

* Firstly, there are changes to make the build of the unit tests more
  flexible and maintainable. These 7 patches switch the unit tests from
  having separate hard-coded lists of files and tests, with lots of
  conditional logic to add/remove items from the list, to instead being
  a single list of files and dependencies per-file. The actual lists of
  test-suites are built up dynamically from the list of files whose
  dependencies are met and therefore will be part of the build.
  This rework enables...
* The second set of patches which expands the list of libraries which
  can be disabled at build-time, while still giving a successful build
  of DPDK. Overall, after this set is applied, the list of optional
  libraries in DPDK is approx doubled in size. [15 additions giving
  31 optional libs]

NOTES:
* as stated in the RFC, the test binary moves in the build as a result
  of this set - from app/test/dpdk-test to app/dpdk-test.
* running the suites of tests no longer works after patch 4 until patch
  5-7 is applied. If this is an issue, the patches can be merged on
  apply. However, the individual tests are still runnable through the
  test binary - it's only through "meson test" that there are issues, so
  I'm keeping the patches separate since each is relatively complex in its
  own right.
* A number of files in the test directory were not being built, and seem
  to have never been built with meson. This requires separate work, outside
  the scope of the patchset to fix.
* Similarly, some unit tests were never added to test suites. These can
  now be found by using grep to find the old REGISTER_TEST_COMMAND macro.
  Again, adding orphan tests to suites is outside the scope of this set.

V5:
- revert back to using the environment to pass the test names rather than
  the commandline. This allows use of "--test-args" with meson test to
  pass additional EAL parameters, such as coremasks, to the test run.
- add back in support for the script for checking telemetry commands
- add support for the dump-* tests in the "debug-tests" suite. Sadly,
  even though these are hardcoded in commands.c, they aren't easily
  discoverable at build-time, so we have to use a hard-coded list.

V4:
- improve test suite definition, hopefully improving test pass rate:
  - drop unnecessary file-prefix for fast tests, since tests are not parallel
  - always pass driver path flag in "meson test" for shared builds.

V3:
- Create symlink to old path to test binary, to ensure all automated CI,
  and any other use of the test binary still works.

V2:
- Changes since RFC:
   -  merged two previous sets - reworking test build, and disabling
  libs - into one, since they depend on each other
   -  Reworked the list of dependencies for each file in the tests,
  so that files that depend on other files have those deps called
  out.
   -  Added list of external dependencies to the test builds
   -  For disabling libs, merged patches disabling individual libs
  together into one patch, since all unit test build changes were
  unnecessary. This made each patch a single-line addition to
  the optional list in lib/meson.build.

Bruce Richardson (10):
  app/test: add new macros for various test types
  app/test: tag tests with the test type
  app/test: make telemetry data test buildable on windows
  app/test: build using per-file dependency matrix
  app/test: define unit tests suites based on test macros
  app/test: add test case for scripted telemetry commands
  app/test: add debug test suite
  examples/l3fwd: make eventdev an optional dependency
  build: make most device classes optional
  build: expand list of optional libraries

 app/meson.build   |  15 +-
 app/test/meson.build  | 823 +-
 app/test/suites/meson.build   | 124 +++
 app/test/{ => suites}/test_telemetry.sh   |   0
 app/test/test.h   |   9 +-
 app/test/test_acl.c   |   2 +-
 app/test/test_atomic.c|   2 +-
 app/test/test_barrier.c   |   2 +-
 app/test/test_bitmap.c|   2 +-
 app/test/test_bitops.c|   2 +-
 app/test/test_bpf.c   |   4 +-
 app/test/test_byteorder.c |   2 +-
 app/test/test_cksum.c |   2 +-
 app/test/test_cmdline.c   |   2 +-
 app/test/test_common.c|   2 +-
 app/test/test_cpuflags.c  |   2 +-
 app/test/test_crc.c   |   2 +-
 app/test/test_cryptodev.c |  38 +-
 app/test/test_cryptodev_asym.c|   2 +-
 app/test/test_cycles.c|   2 +-
 app/test/test_debug.c |   

[PATCH v5 02/10] app/test: tag tests with the test type

[Bruce Richardson]

Rather than having the test types called out in the meson.build file, we
can use macros to identify the test type in the C file itself and then
dynamically build up the tests lists at config time.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 app/test/test_acl.c  |  2 +-
 app/test/test_atomic.c   |  2 +-
 app/test/test_barrier.c  |  2 +-
 app/test/test_bitmap.c   |  2 +-
 app/test/test_bitops.c   |  2 +-
 app/test/test_bpf.c  |  4 +--
 app/test/test_byteorder.c|  2 +-
 app/test/test_cksum.c|  2 +-
 app/test/test_cmdline.c  |  2 +-
 app/test/test_common.c   |  2 +-
 app/test/test_cpuflags.c |  2 +-
 app/test/test_crc.c  |  2 +-
 app/test/test_cryptodev.c| 38 
 app/test/test_cryptodev_asym.c   |  2 +-
 app/test/test_cycles.c   |  2 +-
 app/test/test_debug.c|  2 +-
 app/test/test_devargs.c  |  2 +-
 app/test/test_distributor.c  |  2 +-
 app/test/test_distributor_perf.c |  2 +-
 app/test/test_dmadev.c   |  2 +-
 app/test/test_eal_flags.c| 24 +++
 app/test/test_eal_fs.c   |  2 +-
 app/test/test_efd.c  |  2 +-
 app/test/test_efd_perf.c |  2 +-
 app/test/test_errno.c|  2 +-
 app/test/test_ethdev_link.c  |  2 +-
 app/test/test_event_ring.c   |  2 +-
 app/test/test_eventdev.c |  2 +-
 app/test/test_fbarray.c  |  2 +-
 app/test/test_fib.c  |  4 +--
 app/test/test_fib6.c |  4 +--
 app/test/test_fib6_perf.c|  2 +-
 app/test/test_fib_perf.c |  2 +-
 app/test/test_func_reentrancy.c  |  2 +-
 app/test/test_hash.c |  2 +-
 app/test/test_hash_functions.c   |  2 +-
 app/test/test_hash_multiwriter.c |  2 +-
 app/test/test_hash_perf.c|  2 +-
 app/test/test_hash_readwrite.c   |  4 +--
 app/test/test_hash_readwrite_lf_perf.c   |  2 +-
 app/test/test_interrupts.c   |  2 +-
 app/test/test_ipfrag.c   |  2 +-
 app/test/test_ipsec.c|  2 +-
 app/test/test_ipsec_perf.c   |  2 +-
 app/test/test_kvargs.c   |  2 +-
 app/test/test_lcores.c   |  2 +-
 app/test/test_logs.c |  2 +-
 app/test/test_lpm.c  |  2 +-
 app/test/test_lpm6.c |  2 +-
 app/test/test_lpm6_perf.c|  2 +-
 app/test/test_lpm_perf.c |  2 +-
 app/test/test_malloc.c   |  2 +-
 app/test/test_malloc_perf.c  |  2 +-
 app/test/test_mbuf.c |  2 +-
 app/test/test_mcslock.c  |  2 +-
 app/test/test_member.c   |  2 +-
 app/test/test_member_perf.c  |  2 +-
 app/test/test_memcpy.c   |  2 +-
 app/test/test_memcpy_perf.c  |  2 +-
 app/test/test_memory.c   |  2 +-
 app/test/test_mempool.c  |  2 +-
 app/test/test_mempool_perf.c |  2 +-
 app/test/test_memzone.c  |  2 +-
 app/test/test_meter.c|  2 +-
 app/test/test_mp_secondary.c |  2 +-
 app/test/test_per_lcore.c|  2 +-
 app/test/test_pflock.c   |  2 +-
 app/test/test_pie.c  |  6 ++--
 app/test/test_pmd_perf.c |  2 +-
 app/test/test_power.c|  2 +-
 app/test/test_power_cpufreq.c|  2 +-
 app/test/test_power_intel_uncore.c   |  2 +-
 app/test/test_power_kvm_vm.c |  2 +-
 app/test/test_prefetch.c |  2 +-
 app/test/test_rand_perf.c|  2 +-
 app/test/test_rcu_qsbr.c |  2 +-
 app/test/test_rcu_qsbr_perf.c|  2 +-
 app/test/test_reassembly_perf.c  |  2 +-
 app/test/test_reciprocal_division.c  |  2 +-
 app/test/test_reciprocal_division_perf.c |  2 +-
 app/test/test_red.c  |  4 +--
 app/test/test_reorder.c  |  2 +-
 app/test/test_rib.c  |  4 +--
 app/test/test_rib6.c |  4 +--
 app/test/test_ring.c |  2 +-
 app/test/test_ring_perf.c|  2 +-
 app/test/test_rwlock.c   |  8 ++---
 app/test/test_sched.c|  2 +-
 app/test/test_security.c |  2 +-
 app/test/test_seqlock.c  |  2 +-
 app/test/test_service_cores.c|  4 +--
 app/test/test_spinlock.c |  2 +-
 app/test/test_stack.c|  4 +--
 app/test/test_stack_perf.c 

[PATCH v5 03/10] app/test: make telemetry data test buildable on windows

[Bruce Richardson]

To enable the building of the telemetry data tests file when building on
windows, we need to provide a stub implementation. That way, the test
file is buildable any time the library itself is built.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 app/test/test_telemetry_data.c | 14 --
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/app/test/test_telemetry_data.c b/app/test/test_telemetry_data.c
index a960f44c1b..b2dc1d75db 100644
--- a/app/test/test_telemetry_data.c
+++ b/app/test/test_telemetry_data.c
@@ -2,11 +2,20 @@
  * Copyright 2020 Intel Corporation
  */
 
+#ifdef RTE_EXEC_ENV_WINDOWS
+#include "test.h"
+
+static int
+telemetry_data_autotest(void)
+{
+   return TEST_SKIPPED;
+}
+
+#else
+
 #include 
 #include 
-#ifndef RTE_EXEC_ENV_WINDOWS
 #include 
-#endif
 #include 
 #include 
 
@@ -604,5 +613,6 @@ telemetry_data_autotest(void)
close(sock);
return 0;
 }
+#endif /* windows/non-windows */
 
 REGISTER_TEST_COMMAND(telemetry_data_autotest, telemetry_data_autotest);
-- 
2.39.2

[PATCH v5 04/10] app/test: build using per-file dependency matrix

[Bruce Richardson]

Rather than using if-else constructs to selectively add or remove files
from the UT build, switch to a table-based approach where each file
lists out what libs or drivers it depends upon.

Initial version of this table was generated via analysis of the header
files included in each C file. The basic dependencies of the test binary
[cmdline, ring, mempool and mbuf] were then removed from the per-file
lists, as there is no point in checking them as the whole app will be
disabled if they are not present.

With the file list, the dependencies for the "utility" C-files are kept
separate, so that other tests which depend on the functions provided by
those files can have that dependency recorded properly. The basic
cryptodev tests also fall into this category as functions from the main
cryptodev test file are used by other crypto tests.

As well as the main table for internal dependencies, some test
files have separate external components too. A second, much smaller
table lists these dependencies.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 app/meson.build  |   8 +-
 app/test/meson.build | 823 ---
 2 files changed, 238 insertions(+), 593 deletions(-)

diff --git a/app/meson.build b/app/meson.build
index 4fc1a83eba..0d8b618e7f 100644
--- a/app/meson.build
+++ b/app/meson.build
@@ -32,6 +32,11 @@ apps = [
 'test-security-perf',
 ]
 
+if get_option('tests')
+# build the auto test app if enabled.
+apps += 'test'
+endif
+
 default_cflags = machine_args + ['-DALLOW_EXPERIMENTAL_API']
 default_ldflags = []
 if get_option('default_library') == 'static' and not is_windows
@@ -106,6 +111,3 @@ foreach app:apps
 install_rpath: join_paths(get_option('prefix'), 
driver_install_path),
 install: true)
 endforeach
-
-# special case the autotests
-subdir('test')
diff --git a/app/test/meson.build b/app/test/meson.build
index 66897c14a3..2b885ae273 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -1,446 +1,230 @@
 # SPDX-License-Identifier: BSD-3-Clause
-# Copyright(c) 2017 Intel Corporation
-
-if not get_option('tests')
-subdir_done()
-endif
-
-test_sources = files(
-'commands.c',
-'packet_burst_generator.c',
-'test.c',
-'test_acl.c',
-'test_alarm.c',
-'test_atomic.c',
-'test_barrier.c',
-'test_bitops.c',
-'test_bitmap.c',
-'test_bpf.c',
-'test_byteorder.c',
-'test_cksum.c',
-'test_cksum_perf.c',
-'test_cmdline.c',
-'test_cmdline_cirbuf.c',
-'test_cmdline_etheraddr.c',
-'test_cmdline_ipaddr.c',
-'test_cmdline_lib.c',
-'test_cmdline_num.c',
-'test_cmdline_portlist.c',
-'test_cmdline_string.c',
-'test_common.c',
-'test_cpuflags.c',
-'test_crc.c',
-'test_cryptodev.c',
-'test_cryptodev_asym.c',
-'test_cryptodev_blockcipher.c',
-'test_cryptodev_crosscheck.c',
-'test_cryptodev_security_ipsec.c',
-'test_cryptodev_security_pdcp.c',
-'test_cycles.c',
-'test_debug.c',
-'test_devargs.c',
-'test_distributor.c',
-'test_distributor_perf.c',
-'test_dmadev.c',
-'test_dmadev_api.c',
-'test_eal_flags.c',
-'test_eal_fs.c',
-'test_efd.c',
-'test_efd_perf.c',
-'test_errno.c',
-'test_ethdev_link.c',
-'test_event_crypto_adapter.c',
-'test_event_eth_rx_adapter.c',
-'test_event_ring.c',
-'test_event_timer_adapter.c',
-'test_eventdev.c',
-'test_external_mem.c',
-'test_fbarray.c',
-'test_fib.c',
-'test_fib_perf.c',
-'test_fib6.c',
-'test_fib6_perf.c',
-'test_func_reentrancy.c',
-'test_hash.c',
-'test_hash_functions.c',
-'test_hash_multiwriter.c',
-'test_hash_readwrite.c',
-'test_hash_perf.c',
-'test_hash_readwrite_lf_perf.c',
-'test_interrupts.c',
-'test_ipfrag.c',
-'test_ipsec.c',
-'test_ipsec_sad.c',
-'test_ipsec_perf.c',
-'test_kvargs.c',
-'test_lcores.c',
-'test_logs.c',
-'test_lpm.c',
-'test_lpm6.c',
-'test_lpm6_perf.c',
-'test_lpm_perf.c',
-'test_malloc.c',
-'test_malloc_perf.c',
-'test_mbuf.c',
-'test_member.c',
-'test_member_perf.c',
-'test_memcpy.c',
-'test_memcpy_perf.c',
-'test_memory.c',
-'test_mempool.c',
-'test_mempool_perf.c',
-'test_memzone.c',
-'test_meter.c',
-'test_mcslock.c',
-'test_mp_secondary.c',
-'test_per_lcore.c',
-'test_pflock.c',
-'test_pmd_perf.c',
-'test_power.c',
-'test_power_cpufreq.c',
-'test_power_kvm_vm.c',
-'test_power_intel_uncore.c',
-'test_prefetch.c',
-

[PATCH v5 06/10] app/test: add test case for scripted telemetry commands

[Bruce Richardson]

The test script for calling all telemetry commands is not discoverable
by checking the C files for the build. Therefore we need to add it in as
a special-case test in the fast-tests group.

Signed-off-by: Bruce Richardson 
---
 app/test/suites/meson.build | 30 +
 app/test/{ => suites}/test_telemetry.sh |  0
 2 files changed, 30 insertions(+)
 rename app/test/{ => suites}/test_telemetry.sh (100%)

diff --git a/app/test/suites/meson.build b/app/test/suites/meson.build
index ec74d8adf2..2b428847c6 100644
--- a/app/test/suites/meson.build
+++ b/app/test/suites/meson.build
@@ -72,3 +72,33 @@ foreach suite:test_suites
 endforeach
 endif
 endforeach
+
+# standalone test for telemetry
+if not is_windows and dpdk_conf.has('RTE_LIB_TELEMETRY')
+test_args = [dpdk_test]
+test_args += test_no_huge_args
+if get_option('default_library') == 'shared'
+test_args += ['-d', dpdk_drivers_build_dir]
+endif
+if dpdk_conf.has('RTE_CRYPTO_NULL')
+test_args += ['--vdev=crypto_null0']
+endif
+if dpdk_conf.has('RTE_DMA_SKELETON')
+test_args += ['--vdev=dma_skeleton0']
+endif
+if dpdk_conf.has('RTE_EVENT_SKELETON')
+test_args += ['--vdev=event_skeleton0']
+endif
+if dpdk_conf.has('RTE_NET_NULL')
+test_args += ['--vdev=net_null0']
+endif
+if dpdk_conf.has('RTE_RAW_SKELETON')
+test_args += ['--vdev=rawdev_skeleton0']
+endif
+test_args += ['-a', ':00:00.0']
+test('telemetry_all', find_program('test_telemetry.sh'),
+args: test_args,
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'fast-tests')
+endif
diff --git a/app/test/test_telemetry.sh b/app/test/suites/test_telemetry.sh
similarity index 100%
rename from app/test/test_telemetry.sh
rename to app/test/suites/test_telemetry.sh
-- 
2.39.2

[PATCH v5 07/10] app/test: add debug test suite

[Bruce Richardson]

Add in for all platforms, the suite of tests to dump out the structure
information available. Since the commands are defined in commands.c on
all OS's, do not limit their presence to just the non-windows OS's.

Signed-off-by: Bruce Richardson 
---
 app/test/suites/meson.build | 20 
 1 file changed, 20 insertions(+)

diff --git a/app/test/suites/meson.build b/app/test/suites/meson.build
index 2b428847c6..8cf498ac52 100644
--- a/app/test/suites/meson.build
+++ b/app/test/suites/meson.build
@@ -102,3 +102,23 @@ if not is_windows and dpdk_conf.has('RTE_LIB_TELEMETRY')
 is_parallel : false,
 suite : 'fast-tests')
 endif
+
+# dump tests are defined in commands.c, and not easily extractable
+dump_test_names = [
+'dump_devargs',
+'dump_log_types',
+'dump_malloc_heaps',
+'dump_malloc_stats',
+'dump_mempool',
+'dump_memzone',
+'dump_physmem',
+'dump_ring',
+'dump_struct_sizes',
+]
+foreach arg : dump_test_names
+test(arg, dpdk_test,
+env : ['DPDK_TEST=' + arg],
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'debug-tests')
+endforeach
-- 
2.39.2

[PATCH v5 08/10] examples/l3fwd: make eventdev an optional dependency

[Bruce Richardson]

With l3fwd being a very commonly used example app, and built as part of
our CI tests, we need to ensure it's buildable with just about all
supported DPDK configurations.  To enable l3fwd application to be built
when the eventdev library is disabled, we need to compile in the
eventdev support conditionally.  Thankfully, the eventdev support is
pretty self-contained, with only the main.c file having more than a
couple of ifdefs.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 examples/l3fwd/l3fwd_em.c  |  2 +
 examples/l3fwd/l3fwd_em_hlm.h  |  2 +
 examples/l3fwd/l3fwd_event.c   |  2 +
 examples/l3fwd/l3fwd_event.h   |  7 ++-
 examples/l3fwd/l3fwd_event_generic.c   |  2 +
 examples/l3fwd/l3fwd_event_internal_port.c |  2 +
 examples/l3fwd/l3fwd_fib.c |  2 +
 examples/l3fwd/l3fwd_lpm.c |  2 +
 examples/l3fwd/main.c  | 65 +-
 examples/l3fwd/meson.build |  5 +-
 10 files changed, 74 insertions(+), 17 deletions(-)

diff --git a/examples/l3fwd/l3fwd_em.c b/examples/l3fwd/l3fwd_em.c
index 476ac0c54f..40e102b38a 100644
--- a/examples/l3fwd/l3fwd_em.c
+++ b/examples/l3fwd/l3fwd_em.c
@@ -663,6 +663,7 @@ em_main_loop(__rte_unused void *dummy)
return 0;
 }
 
+#ifdef RTE_LIB_EVENTDEV
 static __rte_always_inline void
 em_event_loop_single(struct l3fwd_event_resources *evt_rsrc,
const uint8_t flags)
@@ -959,6 +960,7 @@ em_event_main_loop_tx_q_burst_vector(__rte_unused void 
*dummy)
em_event_loop_vector(evt_rsrc, L3FWD_EVENT_TX_ENQ);
return 0;
 }
+#endif
 
 /* Initialize exact match (hash) parameters. 8< */
 void
diff --git a/examples/l3fwd/l3fwd_em_hlm.h b/examples/l3fwd/l3fwd_em_hlm.h
index 2e11eefad7..31cda9ddc1 100644
--- a/examples/l3fwd/l3fwd_em_hlm.h
+++ b/examples/l3fwd/l3fwd_em_hlm.h
@@ -255,6 +255,7 @@ l3fwd_em_send_packets(int nb_rx, struct rte_mbuf 
**pkts_burst, uint16_t portid,
send_packets_multi(qconf, pkts_burst, dst_port, nb_rx);
 }
 
+#ifdef RTE_LIB_EVENTDEV
 /*
  * Buffer optimized handling of events, invoked
  * from main_loop.
@@ -347,5 +348,6 @@ l3fwd_em_process_event_vector(struct rte_event_vector *vec,
 
process_event_vector(vec, dst_port);
 }
+#endif /* RTE_LIB_EVENTDEV */
 
 #endif /* __L3FWD_EM_HLM_H__ */
diff --git a/examples/l3fwd/l3fwd_event.c b/examples/l3fwd/l3fwd_event.c
index 32906ab08d..d72a4138cc 100644
--- a/examples/l3fwd/l3fwd_event.c
+++ b/examples/l3fwd/l3fwd_event.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 #include 
 
@@ -341,3 +342,4 @@ l3fwd_event_worker_cleanup(uint8_t event_d_id, uint8_t 
event_p_id,
rte_event_port_quiesce(event_d_id, event_p_id, l3fwd_event_port_flush,
   NULL);
 }
+#endif /* #ifdef RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_event.h b/examples/l3fwd/l3fwd_event.h
index e21817c36b..1fd6fe4a78 100644
--- a/examples/l3fwd/l3fwd_event.h
+++ b/examples/l3fwd/l3fwd_event.h
@@ -6,11 +6,13 @@
 #define __L3FWD_EVENTDEV_H__
 
 #include 
+#include 
+#include 
+
+#ifdef RTE_LIB_EVENTDEV
 #include 
 #include 
 #include 
-#include 
-#include 
 
 #include "l3fwd.h"
 
@@ -164,4 +166,5 @@ void l3fwd_event_worker_cleanup(uint8_t event_d_id, uint8_t 
event_p_id,
struct rte_event events[], uint16_t nb_enq,
uint16_t nb_deq, uint8_t is_vector);
 
+#endif /* ifdef RTE_LIB_EVENTDEV */
 #endif /* __L3FWD_EVENTDEV_H__ */
diff --git a/examples/l3fwd/l3fwd_event_generic.c 
b/examples/l3fwd/l3fwd_event_generic.c
index c80573fc58..ddb6e5c38d 100644
--- a/examples/l3fwd/l3fwd_event_generic.c
+++ b/examples/l3fwd/l3fwd_event_generic.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 
 #include "l3fwd.h"
@@ -309,3 +310,4 @@ l3fwd_event_set_generic_ops(struct l3fwd_event_setup_ops 
*ops)
ops->event_port_setup = l3fwd_event_port_setup_generic;
ops->adapter_setup = l3fwd_rx_tx_adapter_setup_generic;
 }
+#endif /* RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_event_internal_port.c 
b/examples/l3fwd/l3fwd_event_internal_port.c
index 32cf657148..cb49a8b9fa 100644
--- a/examples/l3fwd/l3fwd_event_internal_port.c
+++ b/examples/l3fwd/l3fwd_event_internal_port.c
@@ -2,6 +2,7 @@
  * Copyright(C) 2019 Marvell International Ltd.
  */
 
+#ifdef RTE_LIB_EVENTDEV
 #include 
 
 #include "l3fwd.h"
@@ -311,3 +312,4 @@ l3fwd_event_set_internal_port_ops(struct 
l3fwd_event_setup_ops *ops)
ops->event_port_setup = l3fwd_event_port_setup_internal_port;
ops->adapter_setup = l3fwd_rx_tx_adapter_setup_internal_port;
 }
+#endif /* RTE_LIB_EVENTDEV */
diff --git a/examples/l3fwd/l3fwd_fib.c b/examples/l3fwd/l3fwd_fib.c
index 18398492ae..6a21984415 100644
--- a/examples/l3fwd/l3fwd_fib.c
+++ b/examples/l3fwd/l3fwd_fib.c
@@ -253,6 +253,7

[PATCH v5 05/10] app/test: define unit tests suites based on test macros

[Bruce Richardson]

Rather than having the test suites listed out in the meson.build files
and having to have them enabled/disabled selectively based on what libs
are being built, pull the tests to run from the source files which were
added to the build.

Most test suites require no additional info other than the list of test
names in the suite. However the fast-test are special that they have
additional parameters associated with them. This requires some
additional work in the test extraction script and in the meson.build
file for processing the output.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 app/meson.build   |  7 +-
 app/test/suites/meson.build   | 74 +++
 buildtools/get-test-suites.py | 33 +
 .../has-hugepages.py  |  0
 buildtools/meson.build|  2 +
 5 files changed, 115 insertions(+), 1 deletion(-)
 create mode 100644 app/test/suites/meson.build
 create mode 100644 buildtools/get-test-suites.py
 rename app/test/has_hugepage.py => buildtools/has-hugepages.py (100%)

diff --git a/app/meson.build b/app/meson.build
index 0d8b618e7f..c14dc80892 100644
--- a/app/meson.build
+++ b/app/meson.build
@@ -101,7 +101,7 @@ foreach app:apps
 link_libs = dpdk_static_libraries + dpdk_drivers
 endif
 
-executable('dpdk-' + name,
+exec = executable('dpdk-' + name,
 sources,
 c_args: cflags,
 link_args: ldflags,
@@ -110,4 +110,9 @@ foreach app:apps
 include_directories: includes,
 install_rpath: join_paths(get_option('prefix'), 
driver_install_path),
 install: true)
+if name == 'test'
+dpdk_test = exec
+autotest_sources = sources
+subdir('test/suites')  # define the pre-canned test suites
+endif
 endforeach
diff --git a/app/test/suites/meson.build b/app/test/suites/meson.build
new file mode 100644
index 00..ec74d8adf2
--- /dev/null
+++ b/app/test/suites/meson.build
@@ -0,0 +1,74 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2023 Intel Corporation
+
+# some perf tests (eg: memcpy perf autotest)take very long
+# to complete, so timeout to 10 minutes
+timeout_seconds = 600
+timeout_seconds_fast = 10
+
+test_no_huge_args = ['--no-huge', '-m', '2048']
+has_hugepage = run_command(has_hugepages_cmd, check: true).stdout().strip() != 
'0'
+message('hugepage availability: @0@'.format(has_hugepage))
+
+# process source files to determine the different unit test suites
+# - fast_tests
+# - perf_tests
+# - driver_tests
+test_suites = run_command(get_test_suites_cmd, autotest_sources,
+ check: true).stdout().strip().split()
+foreach suite:test_suites
+# simple cases - tests without parameters or special handling
+suite = suite.split('=')
+suite_name = suite[0]
+suite_tests = suite[1].split(',')
+if suite_name != 'fast-tests'
+# simple cases - tests without parameters or special handling
+foreach t: suite_tests
+test(t, dpdk_test,
+env: ['DPDK_TEST=' + t],
+timeout: timeout_seconds,
+is_parallel: false,
+suite: suite_name)
+endforeach
+else
+# special fast-test handling here
+foreach t: suite_tests
+params = t.split(':')
+test_name = params[0]
+nohuge = params[1] == 'true'
+asan = params[2] == 'true'
+
+test_args = []
+if nohuge
+test_args += test_no_huge_args
+elif not has_hugepage
+continue  #skip this tests
+endif
+if not asan and (get_option('b_sanitize') == 'address'
+or get_option('b_sanitize') == 'address,undefined')
+continue  # skip this test
+endif
+
+if get_option('default_library') == 'shared'
+test_args += ['-d', dpdk_drivers_build_dir]
+endif
+
+test(test_name, dpdk_test,
+args : test_args,
+env: ['DPDK_TEST=' + test_name],
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'fast-tests')
+if not is_windows and test_name == 'trace_autotest'
+test_args += ['--trace=.*']
+test_args += 
['--trace-dir=@0@'.format(meson.current_build_dir())]
+test(test_name + '_with_traces', dpdk_test,
+args : test_args,
+env: ['DPDK_TEST=' + test_name],
+timeout : timeout_seconds_fast,
+is_parallel : false,
+suite : 'fast-tests')
+endif
+endforeach
+endif
+endforeach
diff --git a/buildtools/get-test-suites.py b/buildtools/get-test-suites.py
new file mode 100644
index 00..95a9cad4c8
--- /dev/null

[PATCH v5 09/10] build: make most device classes optional

[Bruce Richardson]

Apart from ethdev and cryptodev, which have lots of components and tests
which depend on them, we can make the device class libraries optional
without too much work.

This patch marks:
* bbdev,
* compressdev,
* dmadev,
* eventdev,
* mldev,
* rawdev,
* regexdev
optional, and ensures that DPDK - including tests - can be built with
these components disabled.

Signed-off-by: Bruce Richardson 
Acked-by: Morten Brørup 
---
 lib/meson.build | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/lib/meson.build b/lib/meson.build
index 92cbd6bb66..1a78c8c984 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -67,8 +67,12 @@ libraries = [
 ]
 
 optional_libs = [
+'bbdev',
 'bitratestats',
 'cfgfile',
+'compressdev',
+'dmadev',
+'eventdev',
 'gpudev',
 'graph',
 'gro',
@@ -76,11 +80,14 @@ optional_libs = [
 'jobstats',
 'latencystats',
 'metrics',
+'mldev',
 'node',
 'pdump',
 'pipeline',
 'port',
 'power',
+'rawdev',
+'regexdev',
 'table',
 'vhost',
 ]
-- 
2.39.2

[PATCH v3 1/2] devtools: add check on symbol maps format

[David Marchand]

Add a check on symbol maps format (fixing ethdev at the same time).
This will be required by a next commit.

Signed-off-by: David Marchand 
---
Changes since v2:
- fixed FreeBSD build by replacing (|pattern) with (pattern)?,

Changes since v1:
- moved this check in a separate patch,
- fixed ethdev map file,

---
 devtools/check-symbol-maps.sh | 21 +
 lib/ethdev/version.map| 30 ++
 2 files changed, 35 insertions(+), 16 deletions(-)

diff --git a/devtools/check-symbol-maps.sh b/devtools/check-symbol-maps.sh
index 8c116bfa9c..1907890635 100755
--- a/devtools/check-symbol-maps.sh
+++ b/devtools/check-symbol-maps.sh
@@ -74,4 +74,25 @@ if [ -n "$empty_maps" ] ; then
 ret=1
 fi
 
+find_bad_format_maps ()
+{
+for map in $@ ; do
+cat $map | awk '
+/^(DPDK_[0-9]*|EXPERIMENTAL|INTERNAL) {$/ { next; } # start of a 
section
+/^};$/ { next; } # end of a section
+/^$/ { next; } # empty line
+/^\t(global:|local: \*;)$/ { next; } # qualifiers
+/^\t[a-zA-Z_0-9]*;( # WINDOWS_NO_EXPORT)?$/ { next; } # symbols
+/^\t# added in [0-9]*\.[0-9]*$/ { next; } # version comments
+{ print $0; }' || echo $map
+done
+}
+
+bad_format_maps=$(find_bad_format_maps $@)
+if [ -n "$bad_format_maps" ] ; then
+echo "Found badly formatted maps:"
+echo "$bad_format_maps"
+ret=1
+fi
+
 exit $ret
diff --git a/lib/ethdev/version.map b/lib/ethdev/version.map
index b965d6aa52..da4f847116 100644
--- a/lib/ethdev/version.map
+++ b/lib/ethdev/version.map
@@ -181,7 +181,20 @@ EXPERIMENTAL {
__rte_ethdev_trace_tx_burst;
rte_flow_get_aged_flows;
 
-   # Marked as experimental in 20.11
+   # added in 20.11
+   rte_eth_hairpin_bind;
+   rte_eth_hairpin_get_peer_ports;
+   rte_eth_hairpin_unbind;
+   rte_eth_link_speed_to_str;
+   rte_eth_link_to_str;
+   rte_eth_fec_get_capability;
+   rte_eth_fec_get;
+   rte_eth_fec_set;
+   rte_flow_tunnel_decap_set;
+   rte_flow_tunnel_match;
+   rte_flow_get_restore_info;
+   rte_flow_tunnel_action_decap_release;
+   rte_flow_tunnel_item_release;
rte_tm_capabilities_get;
rte_tm_get_number_of_leaf_nodes;
rte_tm_hierarchy_commit;
@@ -213,21 +226,6 @@ EXPERIMENTAL {
rte_tm_wred_profile_add;
rte_tm_wred_profile_delete;
 
-   # added in 20.11
-   rte_eth_hairpin_bind;
-   rte_eth_hairpin_get_peer_ports;
-   rte_eth_hairpin_unbind;
-   rte_eth_link_speed_to_str;
-   rte_eth_link_to_str;
-   rte_eth_fec_get_capability;
-   rte_eth_fec_get;
-   rte_eth_fec_set;
-   rte_flow_tunnel_decap_set;
-   rte_flow_tunnel_match;
-   rte_flow_get_restore_info;
-   rte_flow_tunnel_action_decap_release;
-   rte_flow_tunnel_item_release;
-
# added in 21.02
rte_eth_get_monitor_addr;
 
-- 
2.41.0

[PATCH v3 2/2] devtools: list symbols by version

[David Marchand]

Make it easier to list experimental symbols added in a certain version.

Example:
$ ./buildtools/map-list-symbol.sh -V 18.11 lib/eal/version.map
lib/eal/version.map EXPERIMENTAL rte_dev_event_callback_process
lib/eal/version.map EXPERIMENTAL rte_dev_hotplug_handle_disable
lib/eal/version.map EXPERIMENTAL rte_dev_hotplug_handle_enable

Signed-off-by: David Marchand 
Acked-by: Stephen Hemminger 
---
 buildtools/map-list-symbol.sh | 39 ++-
 1 file changed, 25 insertions(+), 14 deletions(-)

diff --git a/buildtools/map-list-symbol.sh b/buildtools/map-list-symbol.sh
index 3bf9bd66f8..a834399816 100755
--- a/buildtools/map-list-symbol.sh
+++ b/buildtools/map-list-symbol.sh
@@ -6,7 +6,7 @@ section=all
 symbol=all
 quiet=
 
-while getopts 'S:s:q' name; do
+while getopts 'S:s:qV:' name; do
case $name in
S)
[ $section = 'all' ] || {
@@ -25,8 +25,11 @@ while getopts 'S:s:q' name; do
q)
quiet='y'
;;
+   V)
+   version=$OPTARG
+   ;;
?)
-   echo 'usage: $0 [-S section] [-s symbol] [-q]'
+   echo 'usage: $0 [-S section] [-s symbol] [-V version] [-q]'
exit 1
;;
esac
@@ -38,7 +41,8 @@ for file in $@; do
cat "$file" |awk '
BEGIN {
current_section = "";
-   if ("'$section'" == "all" && "'$symbol'" == "all") {
+   current_version = "";
+   if ("'$section'" == "all" && "'$symbol'" == "all" && 
"'$version'" == "") {
ret = 0;
} else {
ret = 1;
@@ -49,18 +53,25 @@ for file in $@; do
current_section = $1;
}
}
-   /.*}/ { current_section = ""; }
+   /.*}/ { current_section = ""; current_version = ""; }
+   /^\t# added in / {
+   current_version=$4;
+   }
/^[^}].*[^:*];/ {
-   if (current_section != "") {
-   gsub(";","");
-   if ("'$symbol'" == "all" || $1 == "'$symbol'") {
-   ret = 0;
-   if ("'$quiet'" == "") {
-   print "'$file' "current_section" "$1;
-   }
-   if ("'$symbol'" != "all") {
-   exit 0;
-   }
+   if (current_section == "") {
+   next;
+   }
+   if ("'$version'" != "" && "'$version'" != current_version) {
+   next;
+   }
+   gsub(";","");
+   if ("'$symbol'" == "all" || $1 == "'$symbol'") {
+   ret = 0;
+   if ("'$quiet'" == "") {
+   print "'$file' "current_section" "$1;
+   }
+   if ("'$symbol'" != "all") {
+   exit 0;
}
}
}
-- 
2.41.0

Re: [PATCH v4 0/8] support reinit flow

[Stephen Hemminger]

On Tue, 15 Aug 2023 10:50:15 -0400
ok...@kernel.org wrote:

> From: Sinan Kaya 
> 
> We want to be able to call rte_eal_init() and rte_eal_cleanup()
> APIs back to back for maintanance reasons.
> 
> Here is a summary of the code we have seen so far:
> 
> 1. some code support getting called multiple times by keeping
> a static variable.
> 2. some code initializes once but never clean up after them and
> don't have a cleanup API.
> 3. some code assumes that they only get called once during the
> lifecycle of the process.
> 
> Most changes in this patch center around following the #1 design
> principle.
> 
> Why?
> 
> It is not always ideal to reinitialize a DPDK process. Memory needs
> to be reinitialized, hugetables need to warm up etc.


I am familiar with the backstory of why this is desirable in your case.
But others may not be. It will work for you, but for the wider the
range of libraries and drivers it probably won't.

As a compromise, can this restart be officially tagged as unsupported.
I.e. it may work for some drivers and libraries but not all of them.
If nothing else many parts of DPDK still do leak memory on cleanup
and currently this is harmless.

This has enough impact that it probably needs to wait past 23.11 release.

Could you add a test for restart into standalone tests and test-pmd?

Re: [PATCH v4 8/8] eal: initialize worker threads once

[Stephen Hemminger]

On Tue, 15 Aug 2023 10:50:23 -0400
ok...@kernel.org wrote:

> From: Sinan Kaya 
> 
> Initialize worker threads once and keep a flag
> for other init calls.
> 
> Signed-off-by: Sinan Kaya 
> ---
>  lib/eal/linux/eal.c | 60 -
>  1 file changed, 32 insertions(+), 28 deletions(-)
> 
> diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
> index 584c78e640..1305e1df54 100644
> --- a/lib/eal/linux/eal.c
> +++ b/lib/eal/linux/eal.c
> @@ -80,6 +80,7 @@ int rte_cycles_vmware_tsc_map;
>  
>  
>  static uint32_t run_once;
> +static bool worker_initialized;
>  

I see a pattern here. Many places are using static to
only initialize once. Could you name these variables the
same; suggest using run_once everywhere.

Re: [PATCH v4 2/8] eal: fixes for re-initialization issues

[Stephen Hemminger]

On Tue, 15 Aug 2023 10:50:17 -0400
ok...@kernel.org wrote:

> +static uint32_t run_once;
> +
>  int
>  eal_clean_runtime_dir(void)
>  {
> @@ -505,6 +507,7 @@ eal_parse_socket_arg(char *strval, volatile uint64_t 
> *socket_arg)
>   socket_arg[i] = val;
>   }
>  
> + __atomic_store_n(&run_once, 0, __ATOMIC_RELAXED);
>   return 0;

Interesting, other flags don't use atomic. Why here?

And is already set elsewhere?

Re: [PATCH v4 2/8] eal: fixes for re-initialization issues

[Sinan Kaya]

On Tue, 2023-08-15 at 10:49 -0700, Stephen Hemminger wrote:
> >   
> > + __atomic_store_n(&run_once, 0, __ATOMIC_RELAXED);
> >return 0;
> 
> 
> Interesting, other flags don't use atomic. Why here?
> 
> 
> 
> And is already set elsewhere?

Looking at the history, this variable used to be an atomic
variable. Later, it got replaced with a uint32_t and write
is done with __atomic_store_n instead. I followed other
examples in the same source code file.

Re: [PATCH v4 8/8] eal: initialize worker threads once

[Sinan Kaya]

On Tue, 2023-08-15 at 10:46 -0700, Stephen Hemminger wrote:
> >   static uint32_t run_once;
> > +static bool worker_initialized;
> >   
> 
> 
> I see a pattern here. Many places are using static to
> 
> only initialize once. Could you name these variables the
> 
> same; suggest using run_once everywhere.

sure, will do.

Re: [PATCH v4 0/8] support reinit flow

[Sinan Kaya]

On Tue, 2023-08-15 at 10:45 -0700, Stephen Hemminger wrote:
> > Why?
> > It is not always ideal to reinitialize a DPDK process. Memory needs
> > to be reinitialized, hugetables need to warm up etc.
> 
> 
> 
> 
> I am familiar with the backstory of why this is desirable in your
> case.
> 
> But others may not be. It will work for you, but for the wider the
> 
> range of libraries and drivers it probably won't.
> 
> 

Fair enough.

> 
> As a compromise, can this restart be officially tagged as
> unsupported.

any pointers how to do this?

I have no idea how to mark something unsupported in code.
If this is acceptable in cover letter, I'm happy to do that too.

> 
> I.e. it may work for some drivers and libraries but not all of them.
> 
> If nothing else many parts of DPDK still do leak memory on cleanup
> 
> and currently this is harmless.
> 
> 
> 
> This has enough impact that it probably needs to wait past 23.11
> release.
> 
> 

Sure, no rush. Happy to wait for time slot and accumulate review
feedbacks.

> 
> Could you add a test for restart into standalone tests and test-pmd?

Will do.

Re: [PATCH v5 04/10] app/test: build using per-file dependency matrix

[Patrick Robb]

Adam from our team just raised something important about this patch and UNH
CI which I missed during the RFC discussion.
Presently, eal_flags_file_prefix_autotest fails on arm tx2 systems (arm
people are aware of this, they think it might be a memory leak from mlx5
driver), so at their request we are disabling this unit test (on arm only)
for them for the time being. This failure was discovered when we were
initially standing up arm unit testing on tx2 servers earlier this year -
previously there was no coverage for this. We do this filtering by
going through the /app/test/meson.build file and resetting
the eal_flags_file_prefix_autotest line to an empty string. This process is
broken by your patch series. Again, I'm sorry I didn't catch this concern
when discussing it during the RFC. This is why you are not getting unit
test results for arm64, we can't run unit tests given the changes in this
patch.

Your refactor likely means that going forward, we will no longer be able to
tailor the fast test suite (or any suite) per vendor request. That might
actually be a good thing. In any case, if this is merged now, it is going
to put the tree in a state where our CI doesn't run any unit testing on
ARM. I don't know how close this patch series is to possibly hitting
mainline, but, if possible, can that be delayed, pending us figuring out
how we will respond to this situation?

RE: [EXT] [PATCH v1 0/3] Add support for inter-domain DMA operations

[Satananda Burla]

Hi Anatoly

> -Original Message-
> From: Anatoly Burakov 
> Sent: Friday, August 11, 2023 9:15 AM
> To: dev@dpdk.org
> Cc: bruce.richard...@intel.com
> Subject: [EXT] [PATCH v1 0/3] Add support for inter-domain DMA
> operations
> 
> External Email
> 
> --
> This patchset adds inter-domain DMA operations, and implements driver
> support
> for them in Intel(R) IDXD driver.
> 
> Inter-domain DMA operations are similar to regular DMA operations,
> except that
> source and/or destination addresses will be in virtual address space of
> another
> process. In this patchset, DMA device is extended to support two new
> data plane
> operations: inter-domain copy, and inter-domain fill. No control plane
> API is
> provided for dmadev to set up inter-domain communication (see below for
> more
> info).
Thanks for posting this.
Do you have usecases where a process from 3rd domain sets up transfer 
between memories from 2 domains? i.e process 1 is src, process 2 is
dest and process 3 executes transfer. The SDXI spec also defines this kind
of a transfer.
Have you considered extending  rte_dma_port_param and rte_dma_vchan_conf
to represent interdomain memory transfer setup as a separate port type like
RTE_DMA_PORT_INTER_DOMAIN ?
And then we could have a separate vchan dedicated for this transfer.
The rte_dma_vchan  can be setup with separate struct rte_dma_port_param
each for source and destination. The union could be extended to provide
the necessary information to pmd, this could be set of fields that
would be needed by different architectures like controller id,
pasid, smmu streamid and substreamid etc, if an opaque handle is needed,
it could also be accommodated in the union.
These transfers could also be initiated between 2 processes each having 2
dmadev VFs from the same PF as well. Marvell hardware supports this mode.
Since control plane for this can differ between PMDs, it is better to
setup the memory sharing outside dmadev and only pass the fields of interest to
the PMD for completing the transfer. For instance, for PCIe EP to Host
DMA transactions (MEM_TO_DEV and DEV_TO_MEM), the process of setting up
shared memory from PCIe host is not part of dmadev.
If we wish to make the memory sharing interface as a part of dmadev, then
preferably the control plane has to be abstracted to work for all the modes
and architectures.

Regards
Satananda

Re: [PATCH v4 0/8] support reinit flow

[Stephen Hemminger]

On Tue, 15 Aug 2023 14:58:03 -0400
Sinan Kaya  wrote:

> > 
> > As a compromise, can this restart be officially tagged as
> > unsupported.  
> 
> any pointers how to do this?
> 
> I have no idea how to mark something unsupported in code.
> If this is acceptable in cover letter, I'm happy to do that too.

Put some additional notes in the rte_eal.h like:

diff --git a/lib/eal/include/rte_eal.h b/lib/eal/include/rte_eal.h
index 53c4a5519e61..348b340f006d 100644
--- a/lib/eal/include/rte_eal.h
+++ b/lib/eal/include/rte_eal.h
@@ -67,6 +67,11 @@ int rte_eal_iopl_init(void);
  * as possible in the application's main() function.
  * It puts the WORKER lcores in the WAIT state.
  *
+ * @warning
+ * It maybe possisble to call it again after rte_eal_cleanup().
+ * But this usage is dependent on libraries and drivers support which
+ * may not work. Use at your own risk.
+ *
  * @param argc
  *   A non-negative value.  If it is greater than 0, the array members
  *   for argv[0] through argv[argc] (non-inclusive) shall contain pointers

Or maybe in the Shutdown and Cleanup section of:
 doc/guides/prog_guide/env_abstraction_layer.rst

RE: [PATCH v2] net/ice: write rx timestamp to the first mbuf segment in scattered rx

[Zhang, Qi Z]

> -Original Message-
> From: Martin Weiser 
> Sent: Tuesday, August 8, 2023 7:40 PM
> To: dev@dpdk.org
> Cc: Martin Weiser ; Yang, Qiming
> ; Zhang, Qi Z 
> Subject: [PATCH v2] net/ice: write rx timestamp to the first mbuf segment in
> scattered rx
> 
> Previously, the rx timestamp was written to the last segment of the mbuf 
> chain,
> which was unexpected.

Added:

Fixes: 5543827fc6df ("net/ice: improve performance of Rx timestamp offload")
Cc: sta...@dpdk.org> 

> Signed-off-by: Martin Weiser 
> ---
>  drivers/net/ice/ice_rxtx.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ice/ice_rxtx.c b/drivers/net/ice/ice_rxtx.c index
> 64c4486b4b..e07c6d1f15 100644
> --- a/drivers/net/ice/ice_rxtx.c
> +++ b/drivers/net/ice/ice_rxtx.c
> @@ -2150,7 +2150,7 @@ ice_recv_scattered_pkts(void *rx_queue,
>   }
>   rxq->hw_time_update = rte_get_timer_cycles() /
>(rte_get_timer_hz() / 1000);
> - *RTE_MBUF_DYNFIELD(rxm,
> + *RTE_MBUF_DYNFIELD(first_seg,
>  (ice_timestamp_dynfield_offset),
>  rte_mbuf_timestamp_t *) = ts_ns;
>   pkt_flags |= ice_timestamp_dynflag;
> --
> 2.34.1

Thanks for capture this.

Acked-by: Qi Zhang 

Applied to dpdk-next-net-intel.

RE: [PATCH] bus/pci: fix legacy device IO port map in secondary process

[Ma, WenwuX]

Hi,

> -Original Message-
> From: Gupta, Nipun 
> Sent: 2023年8月13日 14:15
> To: Ma, WenwuX ; dev@dpdk.org
> Cc: david.march...@redhat.com; maxime.coque...@redhat.com; Xia,
> Chenbo ; Li, Miao ; Ling, WeiX
> ; sta...@dpdk.org
> Subject: Re: [PATCH] bus/pci: fix legacy device IO port map in secondary
> process
> 
> Hi Wenwu,
> 
> On 8/7/2023 7:28 AM, Wenwu Ma wrote:
> > When doing IO port mapping for legacy device in secondary process, the
> > region information is missing, so, we need to refill it.
> >
> > Fixes: 4b741542ecde ("bus/pci: avoid depending on private kernel
> > value")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Wenwu Ma 
> > ---
> >   drivers/bus/pci/linux/pci_vfio.c | 21 +
> >   1 file changed, 21 insertions(+)
> >
> > diff --git a/drivers/bus/pci/linux/pci_vfio.c
> > b/drivers/bus/pci/linux/pci_vfio.c
> > index e634de8322..eea1c9851e 100644
> > --- a/drivers/bus/pci/linux/pci_vfio.c
> > +++ b/drivers/bus/pci/linux/pci_vfio.c
> > @@ -1306,6 +1306,11 @@ int
> >   pci_vfio_ioport_map(struct rte_pci_device *dev, int bar,
> > struct rte_pci_ioport *p)
> >   {
> > +   struct vfio_device_info device_info = { .argsz = sizeof(device_info) };
> > +   char pci_addr[PATH_MAX] = {0};
> > +   int vfio_dev_fd;
> > +   struct rte_pci_addr *loc = &dev->addr;
> > +   int ret;
> > uint64_t size, offset;
> >
> > if (bar < VFIO_PCI_BAR0_REGION_INDEX || @@ -1314,6 +1319,22
> @@
> > pci_vfio_ioport_map(struct rte_pci_device *dev, int bar,
> > return -1;
> > }
> >
> > +   if (rte_eal_process_type() == RTE_PROC_SECONDARY) {
> > +   /* store PCI address string */
> > +   snprintf(pci_addr, sizeof(pci_addr), PCI_PRI_FMT,
> > +   loc->domain, loc->bus, loc->devid, loc-
> >function);
> > +
> > +   ret = rte_vfio_setup_device(rte_pci_get_sysfs_path(),
> pci_addr,
> > +   &vfio_dev_fd, &device_info);
> 
> This looks better than fixing in "virtio_remap_pci".
> 
> Ideally, these API should be called irrespective of PRIMARY or SECONDARY
> process here. Miao Li mentioned earlier that "rte_pci_map_device" is called
> from primary process, is it via "virtio_read_caps" API? Isn't there any other
> way to detect if it is a virtio legacy device without calling
> "rte_pci_map_device"?
> 
1. yes, it is via "virtio_read_caps".

2. if (pdev->region[bar].size == 0 && pdev->region[bar].offset == 0)
Do you think it's better this way?

> 
> > +   if (ret)
> > +   return -1;
> > +
> > +   ret = pci_vfio_fill_regions(dev, vfio_dev_fd, &device_info);
> > +   if (ret)
> > +   return -1;
> 
> Corresponding cleanup required in "pci_vfio_ioport_unmap"?
> 
Yes, we need release group fd, I will add it.

Thanks.

> Thanks,
> Nipun
> 
> > +
> > +   }
> > +
> > if (pci_vfio_get_region(dev, bar, &size, &offset) != 0) {
> > RTE_LOG(ERR, EAL, "Cannot get offset of region %d.\n", bar);
> > return -1;

RE: [PATCH] app/dma-perf: fix physical address seg-fault

[Anoob Joseph]

Hi Vipin,

Would using 'rte_pktmbuf_mtod' in 'do_cpu_mem_copy' function resolve the issue?

Thanks,
Anoob

> -Original Message-
> From: Varghese, Vipin 
> Sent: Tuesday, August 15, 2023 8:45 PM
> To: Varghese, Vipin ; tho...@monjalon.net;
> dev@dpdk.org
> Cc: Yigit, Ferruh ; cheng1.ji...@intel.com;
> sta...@dpdk.org
> Subject: [EXT] RE: [PATCH] app/dma-perf: fix physical address seg-fault
> 
> External Email
> 
> --
> [AMD Official Use Only - General]
> 
> Apologies, marking this as `NA`. After recheck of this logic without use of `
> rte_mbuf_data_iova` will result in mbuf meta-data corruption.
> 
> Need to fix this in a different way.
> 
> > -Original Message-
> > From: Vipin Varghese 
> > Sent: Tuesday, August 15, 2023 8:41 PM
> > To: tho...@monjalon.net; dev@dpdk.org
> > Cc: Yigit, Ferruh ; cheng1.ji...@intel.com;
> > sta...@dpdk.org
> > Subject: [PATCH] app/dma-perf: fix physical address seg-fault
> >
> > Caution: This message originated from an External Source. Use proper
> > caution when opening attachments, clicking links, or responding.
> >
> >
> > do_cpu_mem_copy uses DPDK API rte_mbuf_data_iova to return the start
> > of the virtual address for both src and dst.
> > But in case of iova mode set as PA, this results in seg-fault.
> > This is because rte_memcpy VA address and not PA.
> >
> > this fix checks the iova mode and invokes rte_memcpy with the right
> > arguments.
> >
> > Bugzilla ID: 1269
> > Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
> > Cc: cheng1.ji...@intel.com
> >
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Vipin Varghese 
> > ---
> >
> > tested for both va and pa
> >
> > CMD:
> > PA: dpdk-test-dma-perf --iova-mode=pa  -- --config test.ini
> > VA: dpdk-test-dma-perf --iova-mode=va  -- --config test.ini
> > DC: dpdk-test-dma-perf --iova-mode=dc  -- --config test.ini
> >
> > Log: fails for dc mode `EAL: invalid parameters for --iova-mode`
> >
> > test.ini:
> > ```
> > [case1]
> > type=CPU_MEM_COPY
> > mem_size=10
> > buf_size=64,8192,2,MUL
> > src_numa_node=0
> > dst_numa_node=0
> > cache_flush=0
> > test_seconds=2
> > lcore = 7
> > eal_args=--in-memory --no-pci
> > ```
> > ---
> >  app/test-dma-perf/benchmark.c | 36
> > ++-
> >  1 file changed, 27 insertions(+), 9 deletions(-)
> >
> > diff --git a/app/test-dma-perf/benchmark.c
> > b/app/test-dma-perf/benchmark.c index 0601e0d171..73200e1935 100644
> > --- a/app/test-dma-perf/benchmark.c
> > +++ b/app/test-dma-perf/benchmark.c
> > @@ -279,6 +279,10 @@ do_cpu_mem_copy(void *p)
> > struct rte_mbuf **srcs = para->srcs;
> > struct rte_mbuf **dsts = para->dsts;
> > uint32_t i;
> > +   bool isAddrPaMode = false;
> > +
> > +   if (rte_eal_iova_mode() == RTE_IOVA_PA)
> > +   isAddrPaMode = true;
> >
> > worker_info->stop_flag = false;
> > worker_info->ready_flag = true; @@ -286,16 +290,30 @@
> > do_cpu_mem_copy(void *p)
> > while (!worker_info->start_flag)
> > ;
> >
> > -   while (1) {
> > -   for (i = 0; i < nr_buf; i++) {
> > -   /* copy buffer form src to dst */
> > -   rte_memcpy((void 
> > *)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
> > -   (void 
> > *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
> > -   (size_t)buf_size);
> > -   worker_info->total_cpl++;
> > +   if (true == isAddrPaMode) {
> > +   while (1) {
> > +   for (i = 0; i < nr_buf; i++) {
> > +   /* copy buffer form src to dst */
> > +   rte_memcpy((void *)dsts[i],
> > +   (void *)srcs[i],
> > +   (size_t)buf_size);
> > +   worker_info->total_cpl++;
> > +   }
> > +   if (worker_info->stop_flag)
> > +   break;
> > +   }
> > +   } else {
> > +   while (1) {
> > +   for (i = 0; i < nr_buf; i++) {
> > +   /* copy buffer form src to dst */
> > +   rte_memcpy((void
> > *)(uintptr_t)rte_mbuf_data_iova(dsts[i]),
> > +   (void 
> > *)(uintptr_t)rte_mbuf_data_iova(srcs[i]),
> > +   (size_t)buf_size);
> > +   worker_info->total_cpl++;
> > +   }
> > +   if (worker_info->stop_flag)
> > +   break;
> > }
> > -   if (worker_info->stop_flag)
> > -   break;
> > }
> >
> > return 0;
> > --
> > 2.34.1

[PATCH v5 0/2] replace master/slave to main/member

[Chaoyong He]

This patch series aims to replace all the master/slave to main/member
in the bonding PMD. 

---
v5:
* Rebase to DPDK v23.11.
v4:
* Fix the doc compile problem.
v3:
* Fix the check warning about 'CamelCase'.
v2:
* Modify related doc.
* Add 'RTE_DEPRECATED' to related APIs.
---

Long Wu (2):
  ethdev: remove bonding deprecated macro
  net/bonding: replace master/slave to main/member

 app/test-pmd/testpmd.c|  113 +-
 app/test-pmd/testpmd.h|8 +-
 app/test/test_link_bonding.c  | 2792 +
 app/test/test_link_bonding_mode4.c|  588 ++--
 app/test/test_link_bonding_rssconf.c  |  166 +-
 doc/guides/howto/lm_bond_virtio_sriov.rst |   24 +-
 doc/guides/nics/bnxt.rst  |4 +-
 doc/guides/prog_guide/img/bond-mode-1.svg |2 +-
 .../link_bonding_poll_mode_drv_lib.rst|  230 +-
 doc/guides/rel_notes/deprecation.rst  |   19 -
 doc/guides/rel_notes/release_23_11.rst|   19 +
 drivers/net/bonding/bonding_testpmd.c |  178 +-
 drivers/net/bonding/eth_bond_8023ad_private.h |   40 +-
 drivers/net/bonding/eth_bond_private.h|  108 +-
 drivers/net/bonding/rte_eth_bond.h|   96 +-
 drivers/net/bonding/rte_eth_bond_8023ad.c |  372 +--
 drivers/net/bonding/rte_eth_bond_8023ad.h |   67 +-
 drivers/net/bonding/rte_eth_bond_alb.c|   44 +-
 drivers/net/bonding/rte_eth_bond_alb.h|   20 +-
 drivers/net/bonding/rte_eth_bond_api.c|  482 +--
 drivers/net/bonding/rte_eth_bond_args.c   |   32 +-
 drivers/net/bonding/rte_eth_bond_flow.c   |   54 +-
 drivers/net/bonding/rte_eth_bond_pmd.c| 1384 
 drivers/net/bonding/version.map   |   15 +-
 examples/bond/main.c  |   40 +-
 lib/ethdev/rte_ethdev.h   |1 -
 26 files changed, 3488 insertions(+), 3410 deletions(-)

-- 
2.39.1

[PATCH v5 1/2] ethdev: remove bonding deprecated macro

[Chaoyong He]

From: Long Wu 

The macro RTE_ETH_DEV_BONDED_SLAVE was marked as deprecated in v23.07,
we can remove it.

Signed-off-by: Long Wu 
Reviewed-by: Chaoyong He 
---
 doc/guides/rel_notes/deprecation.rst   | 5 +
 doc/guides/rel_notes/release_23_11.rst | 2 ++
 lib/ethdev/rte_ethdev.h| 1 -
 3 files changed, 3 insertions(+), 5 deletions(-)

diff --git a/doc/guides/rel_notes/deprecation.rst 
b/doc/guides/rel_notes/deprecation.rst
index 317875c505..82455f9e18 100644
--- a/doc/guides/rel_notes/deprecation.rst
+++ b/doc/guides/rel_notes/deprecation.rst
@@ -124,10 +124,7 @@ Deprecation Notices
   The legacy actions should be removed
   once ``MODIFY_FIELD`` alternative is implemented in drivers.
 
-* bonding: The macro ``RTE_ETH_DEV_BONDED_SLAVE`` will be
-  deprecated in DPDK 23.07, and removed in DPDK 23.11.
-  The relevant code can be updated using ``RTE_ETH_DEV_BONDING_MEMBER``.
-  The data structure ``struct rte_eth_bond_8023ad_slave_info`` will be
+* bonding: The data structure ``struct rte_eth_bond_8023ad_slave_info`` will be
   renamed to ``struct rte_eth_bond_8023ad_member_info`` in DPDK 23.11.
   The following functions will be removed in DPDK 23.11.
   The old functions:
diff --git a/doc/guides/rel_notes/release_23_11.rst 
b/doc/guides/rel_notes/release_23_11.rst
index 4411bb32c1..2fae9539e2 100644
--- a/doc/guides/rel_notes/release_23_11.rst
+++ b/doc/guides/rel_notes/release_23_11.rst
@@ -87,6 +87,8 @@ Removed Items
 
 * eal: Removed deprecated ``RTE_FUNC_PTR_OR_*`` macros.
 
+* ethdev: Removed deprecated macro ``RTE_ETH_DEV_BONDED_SLAVE``.
+
 * flow_classify: Removed flow classification library and examples.
 
 * kni: Removed the Kernel Network Interface (KNI) library and driver.
diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
index 04a2564f22..a6bf7c16cf 100644
--- a/lib/ethdev/rte_ethdev.h
+++ b/lib/ethdev/rte_ethdev.h
@@ -2038,7 +2038,6 @@ struct rte_eth_dev_owner {
 #define RTE_ETH_DEV_INTR_LSC  RTE_BIT32(1)
 /** Device is a bonding member */
 #define RTE_ETH_DEV_BONDING_MEMBERRTE_BIT32(2)
-#define RTE_ETH_DEV_BONDED_SLAVE RTE_DEPRECATED(RTE_ETH_DEV_BONDED_SLAVE) 
RTE_ETH_DEV_BONDING_MEMBER
 /** Device supports device removal interrupt */
 #define RTE_ETH_DEV_INTR_RMV  RTE_BIT32(3)
 /** Device is port representor */
-- 
2.39.1

[PATCH v2 00/12] net/cpfl: support port representor

[beilei . xing]

From: Beilei Xing 

1. code refine for representor support
2. support port representor

v2 changes:
 - Remove representor data path.
 - Fix coding style.

Beilei Xing (12):
  net/cpfl: refine devargs parse and process
  net/cpfl: introduce interface structure
  net/cpfl: add cp channel
  net/cpfl: enable vport mapping
  net/cpfl: parse representor devargs
  net/cpfl: support probe again
  net/cpfl: create port representor
  net/cpfl: support vport list/info get
  net/cpfl: update vport info before creating representor
  net/cpfl: refine handle virtual channel message
  net/cpfl: support link update for representor
  net/cpfl: support Rx/Tx queue setup for representor

 doc/guides/nics/cpfl.rst   |  36 ++
 doc/guides/rel_notes/release_23_11.rst |   3 +
 drivers/net/cpfl/cpfl_cpchnl.h | 321 +
 drivers/net/cpfl/cpfl_ethdev.c | 619 +
 drivers/net/cpfl/cpfl_ethdev.h |  92 +++-
 drivers/net/cpfl/cpfl_representor.c| 619 +
 drivers/net/cpfl/cpfl_representor.h|  26 ++
 drivers/net/cpfl/cpfl_vchnl.c  |  72 +++
 drivers/net/cpfl/meson.build   |   4 +-
 9 files changed, 1686 insertions(+), 106 deletions(-)
 create mode 100644 drivers/net/cpfl/cpfl_cpchnl.h
 create mode 100644 drivers/net/cpfl/cpfl_representor.c
 create mode 100644 drivers/net/cpfl/cpfl_representor.h
 create mode 100644 drivers/net/cpfl/cpfl_vchnl.c

-- 
2.34.1

[PATCH v2 01/12] net/cpfl: refine devargs parse and process

[beilei . xing]

From: Beilei Xing 

1. Keep devargs in adapter.
2. Refine handling the case with no vport be specified in devargs.
3. Separate devargs parse and devargs process

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c | 154 ++---
 drivers/net/cpfl/cpfl_ethdev.h |   1 +
 2 files changed, 84 insertions(+), 71 deletions(-)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index c4ca9343c3..46b3a52e49 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1407,12 +1407,12 @@ parse_bool(const char *key, const char *value, void 
*args)
 }
 
 static int
-cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter,
-  struct cpfl_devargs *cpfl_args)
+cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
 {
struct rte_devargs *devargs = pci_dev->device.devargs;
+   struct cpfl_devargs *cpfl_args = &adapter->devargs;
struct rte_kvargs *kvlist;
-   int i, ret;
+   int ret;
 
cpfl_args->req_vport_nb = 0;
 
@@ -1445,31 +1445,6 @@ cpfl_parse_devargs(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *adap
if (ret != 0)
goto fail;
 
-   /* check parsed devargs */
-   if (adapter->cur_vport_nb + cpfl_args->req_vport_nb >
-   adapter->max_vport_nb) {
-   PMD_INIT_LOG(ERR, "Total vport number can't be > %d",
-adapter->max_vport_nb);
-   ret = -EINVAL;
-   goto fail;
-   }
-
-   for (i = 0; i < cpfl_args->req_vport_nb; i++) {
-   if (cpfl_args->req_vports[i] > adapter->max_vport_nb - 1) {
-   PMD_INIT_LOG(ERR, "Invalid vport id %d, it should be 0 
~ %d",
-cpfl_args->req_vports[i], 
adapter->max_vport_nb - 1);
-   ret = -EINVAL;
-   goto fail;
-   }
-
-   if (adapter->cur_vports & RTE_BIT32(cpfl_args->req_vports[i])) {
-   PMD_INIT_LOG(ERR, "Vport %d has been requested",
-cpfl_args->req_vports[i]);
-   ret = -EINVAL;
-   goto fail;
-   }
-   }
-
 fail:
rte_kvargs_free(kvlist);
return ret;
@@ -1915,15 +1890,79 @@ cpfl_adapter_ext_deinit(struct cpfl_adapter_ext 
*adapter)
adapter->vports = NULL;
 }
 
+static int
+cpfl_vport_devargs_process(struct cpfl_adapter_ext *adapter)
+{
+   struct cpfl_devargs *devargs = &adapter->devargs;
+   int i;
+
+   /* refine vport number, at least 1 vport */
+   if (devargs->req_vport_nb == 0) {
+   devargs->req_vport_nb = 1;
+   devargs->req_vports[0] = 0;
+   }
+
+   /* check parsed devargs */
+   if (adapter->cur_vport_nb + devargs->req_vport_nb >
+   adapter->max_vport_nb) {
+   PMD_INIT_LOG(ERR, "Total vport number can't be > %d",
+adapter->max_vport_nb);
+   return -EINVAL;
+   }
+
+   for (i = 0; i < devargs->req_vport_nb; i++) {
+   if (devargs->req_vports[i] > adapter->max_vport_nb - 1) {
+   PMD_INIT_LOG(ERR, "Invalid vport id %d, it should be 0 
~ %d",
+devargs->req_vports[i], 
adapter->max_vport_nb - 1);
+   return -EINVAL;
+   }
+
+   if (adapter->cur_vports & RTE_BIT32(devargs->req_vports[i])) {
+   PMD_INIT_LOG(ERR, "Vport %d has been requested",
+devargs->req_vports[i]);
+   return -EINVAL;
+   }
+   }
+
+   return 0;
+}
+
+static int
+cpfl_vport_create(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
+{
+   struct cpfl_vport_param vport_param;
+   char name[RTE_ETH_NAME_MAX_LEN];
+   int ret, i;
+
+   for (i = 0; i < adapter->devargs.req_vport_nb; i++) {
+   vport_param.adapter = adapter;
+   vport_param.devarg_id = adapter->devargs.req_vports[i];
+   vport_param.idx = cpfl_vport_idx_alloc(adapter);
+   if (vport_param.idx == CPFL_INVALID_VPORT_IDX) {
+   PMD_INIT_LOG(ERR, "No space for vport %u", 
vport_param.devarg_id);
+   break;
+   }
+   snprintf(name, sizeof(name), "net_%s_vport_%d",
+pci_dev->device.name,
+adapter->devargs.req_vports[i]);
+   ret = rte_eth_dev_create(&pci_dev->device, name,
+   sizeof(struct cpfl_vport),
+   NULL, NULL, cpfl_dev_vport_init,
+   &vport_param);
+   if (ret != 0)
+   P

[PATCH v2 03/12] net/cpfl: add cp channel

[beilei . xing]

From: Beilei Xing 

Add cpchnl header file.

Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_cpchnl.h | 321 +
 1 file changed, 321 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_cpchnl.h

diff --git a/drivers/net/cpfl/cpfl_cpchnl.h b/drivers/net/cpfl/cpfl_cpchnl.h
new file mode 100644
index 00..c95fad57b6
--- /dev/null
+++ b/drivers/net/cpfl/cpfl_cpchnl.h
@@ -0,0 +1,321 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#ifndef _CPFL_CPCHNL_H_
+#define _CPFL_CPCHNL_H_
+
+/** @brief  Command Opcodes
+ *  Values are to be different from virtchnl.h opcodes
+ */
+enum cpchnl2_ops {
+   /* vport info */
+   CPCHNL2_OP_GET_VPORT_LIST   = 0x8025,
+   CPCHNL2_OP_GET_VPORT_INFO   = 0x8026,
+
+   /* DPHMA Event notifications */
+   CPCHNL2_OP_EVENT= 0x8050,
+};
+
+/* Note! This affects the size of structs below */
+#define CPCHNL2_MAX_TC_AMOUNT  8
+
+#define CPCHNL2_ETH_LENGTH_OF_ADDRESS  6
+
+#define CPCHNL2_FUNC_TYPE_PF   0
+#define CPCHNL2_FUNC_TYPE_SRIOV1
+
+/* vport statuses - must match the DB ones - see enum cp_vport_status*/
+#define CPCHNL2_VPORT_STATUS_CREATED   0
+#define CPCHNL2_VPORT_STATUS_ENABLED   1
+#define CPCHNL2_VPORT_STATUS_DISABLED  2
+#define CPCHNL2_VPORT_STATUS_DESTROYED 3
+
+/* Queue Groups Extension */
+/**/
+
+#define MAX_Q_REGIONS 16
+/* TBD - with current structure sizes, in order not to exceed 4KB ICQH buffer
+ * no more than 11 queue groups are allowed per a single vport..
+ * More will be possible only with future msg fragmentation.
+ */
+#define MAX_Q_VPORT_GROUPS 11
+
+struct cpchnl2_queue_chunk {
+   u32 type;  /* 0:QUEUE_TYPE_TX, 1:QUEUE_TYPE_RX */ /* enum 
nsl_lan_queue_type */
+   u32 start_queue_id;
+   u32 num_queues;
+   u8 pad[4];
+};
+
+/* structure to specify several chunks of contiguous queues */
+struct cpchnl2_queue_grp_chunks {
+   u16 num_chunks;
+   u8 reserved[6];
+   struct cpchnl2_queue_chunk chunks[MAX_Q_REGIONS];
+};
+
+struct cpchnl2_rx_queue_group_info {
+   /* User can ask to update rss_lut size originally allocated
+* by CreateVport command. New size will be returned if allocation 
succeeded,
+* otherwise original rss_size from CreateVport will be returned.
+*/
+   u16 rss_lut_size;
+   u8 pad[6]; /*Future extension purpose*/
+};
+
+struct cpchnl2_tx_queue_group_info {
+   u8 tx_tc; /*TX TC queue group will be connected to*/
+   /* Each group can have its own priority, value 0-7, while each group 
with unique
+* priority is strict priority. It can be single set of queue groups 
which configured with
+* same priority, then they are assumed part of WFQ arbitration group 
and are expected to be
+* assigned with weight.
+*/
+   u8 priority;
+   /* Determines if queue group is expected to be Strict Priority 
according to its priority */
+   u8 is_sp;
+   u8 pad;
+   /* Peak Info Rate Weight in case Queue Group is part of WFQ arbitration 
set.
+* The weights of the groups are independent of each other. Possible 
values: 1-200.
+*/
+   u16 pir_weight;
+   /* Future extension purpose for CIR only */
+   u8 cir_pad[2];
+   u8 pad2[8]; /* Future extension purpose*/
+};
+
+struct cpchnl2_queue_group_id {
+   /* Queue group ID - depended on it's type:
+* Data & p2p - is an index which is relative to Vport.
+* Config & Mailbox - is an ID which is relative to func.
+* This ID is used in future calls, i.e. delete.
+* Requested by host and assigned by Control plane.
+*/
+   u16 queue_group_id;
+   /* Functional type: see CPCHNL2_QUEUE_GROUP_TYPE definitions */
+   u16 queue_group_type;
+   u8 pad[4];
+};
+
+struct cpchnl2_queue_group_info {
+   /* IN */
+   struct cpchnl2_queue_group_id qg_id;
+
+   /* IN, Number of queues of different types in the group. */
+   u16 num_tx_q;
+   u16 num_tx_complq;
+   u16 num_rx_q;
+   u16 num_rx_bufq;
+
+   struct cpchnl2_tx_queue_group_info tx_q_grp_info;
+   struct cpchnl2_rx_queue_group_info rx_q_grp_info;
+
+   u8 egress_port;
+   u8 pad[39]; /*Future extension purpose*/
+   struct cpchnl2_queue_grp_chunks chunks;
+};
+
+struct cpchnl2_queue_groups {
+   u16 num_queue_groups; /* Number of queue groups in struct below */
+   u8 pad[6];
+   /* group information , number is determined by param above */
+   struct cpchnl2_queue_group_info groups[MAX_Q_VPORT_GROUPS];
+};
+
+/**
+ * @brief function types
+ */
+enum cpchnl2_func_type {
+   CPCHNL2_FTYPE_LAN_PF = 0,
+   CPCHNL2_FTYPE_LAN_VF = 1,
+   CPCHNL2_FTYPE_LAN_MAX
+};
+
+/**
+ * @brief containing vport id & type
+ */
+struct cpch

[PATCH v2 02/12] net/cpfl: introduce interface structure

[beilei . xing]

From: Beilei Xing 

Introduce cplf interface structure to distinguish vport and port
representor.

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c |  3 +++
 drivers/net/cpfl/cpfl_ethdev.h | 16 
 2 files changed, 19 insertions(+)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 46b3a52e49..92fe92c00f 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1803,6 +1803,9 @@ cpfl_dev_vport_init(struct rte_eth_dev *dev, void 
*init_params)
goto err;
}
 
+   cpfl_vport->itf.type = CPFL_ITF_TYPE_VPORT;
+   cpfl_vport->itf.adapter = adapter;
+   cpfl_vport->itf.data = dev->data;
adapter->vports[param->idx] = cpfl_vport;
adapter->cur_vports |= RTE_BIT32(param->devarg_id);
adapter->cur_vport_nb++;
diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index b637bf2e45..53e45035e8 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -86,7 +86,19 @@ struct p2p_queue_chunks_info {
uint32_t rx_buf_qtail_spacing;
 };
 
+enum cpfl_itf_type {
+   CPFL_ITF_TYPE_VPORT,
+   CPFL_ITF_TYPE_REPRESENTOR
+};
+
+struct cpfl_itf {
+   enum cpfl_itf_type type;
+   struct cpfl_adapter_ext *adapter;
+   void *data;
+};
+
 struct cpfl_vport {
+   struct cpfl_itf itf;
struct idpf_vport base;
struct p2p_queue_chunks_info *p2p_q_chunks_info;
 
@@ -124,5 +136,9 @@ TAILQ_HEAD(cpfl_adapter_list, cpfl_adapter_ext);
RTE_DEV_TO_PCI((eth_dev)->device)
 #define CPFL_ADAPTER_TO_EXT(p) \
container_of((p), struct cpfl_adapter_ext, base)
+#define CPFL_DEV_TO_VPORT(dev) \
+   ((struct cpfl_vport *)((dev)->data->dev_private))
+#define CPFL_DEV_TO_ITF(dev)   \
+   ((struct cpfl_itf *)((dev)->data->dev_private))
 
 #endif /* _CPFL_ETHDEV_H_ */
-- 
2.34.1

[PATCH v2 04/12] net/cpfl: enable vport mapping

[beilei . xing]

From: Beilei Xing 

1. Handle cpchnl event for vport create/destroy
2. Use hash table to store vport_id to vport_info mapping
3. Use spinlock for thread safe.

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c | 157 +
 drivers/net/cpfl/cpfl_ethdev.h |  21 -
 drivers/net/cpfl/meson.build   |   2 +-
 3 files changed, 177 insertions(+), 3 deletions(-)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 92fe92c00f..17a69c16fe 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -10,6 +10,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include "cpfl_ethdev.h"
 #include "cpfl_rxtx.h"
@@ -1492,6 +1493,108 @@ cpfl_handle_event_msg(struct idpf_vport *vport, uint8_t 
*msg, uint16_t msglen)
}
 }
 
+static int
+cpfl_vport_info_create(struct cpfl_adapter_ext *adapter,
+  struct cpfl_vport_id *vport_identity,
+  struct cpchnl2_vport_info *vport_info)
+{
+   struct cpfl_vport_info *info = NULL;
+   int ret;
+
+   rte_spinlock_lock(&adapter->vport_map_lock);
+   ret = rte_hash_lookup_data(adapter->vport_map_hash, vport_identity, 
(void **)&info);
+   if (ret >= 0) {
+   PMD_DRV_LOG(WARNING, "vport already exist, overwrite info 
anyway");
+   /* overwrite info */
+   if (info)
+   info->vport_info = *vport_info;
+   goto fini;
+   }
+
+   info = rte_zmalloc(NULL, sizeof(*info), 0);
+   if (info == NULL) {
+   PMD_DRV_LOG(ERR, "Failed to alloc memory for vport map info");
+   ret = -ENOMEM;
+   goto err;
+   }
+
+   info->vport_info = *vport_info;
+
+   ret = rte_hash_add_key_data(adapter->vport_map_hash, vport_identity, 
info);
+   if (ret < 0) {
+   PMD_DRV_LOG(ERR, "Failed to add vport map into hash");
+   rte_free(info);
+   goto err;
+   }
+
+fini:
+   rte_spinlock_unlock(&adapter->vport_map_lock);
+   return 0;
+err:
+   rte_spinlock_unlock(&adapter->vport_map_lock);
+   return ret;
+}
+
+static int
+cpfl_vport_info_destroy(struct cpfl_adapter_ext *adapter, struct cpfl_vport_id 
*vport_identity)
+{
+   struct cpfl_vport_info *info;
+   int ret;
+
+   rte_spinlock_lock(&adapter->vport_map_lock);
+   ret = rte_hash_lookup_data(adapter->vport_map_hash, vport_identity, 
(void **)&info);
+   if (ret < 0) {
+   PMD_DRV_LOG(ERR, "vport id not exist");
+   goto err;
+   }
+
+   rte_hash_del_key(adapter->vport_map_hash, vport_identity);
+   rte_spinlock_unlock(&adapter->vport_map_lock);
+   rte_free(info);
+
+   return 0;
+
+err:
+   rte_spinlock_unlock(&adapter->vport_map_lock);
+   return ret;
+}
+
+static void
+cpfl_handle_cpchnl_event_msg(struct cpfl_adapter_ext *adapter, uint8_t *msg, 
uint16_t msglen)
+{
+   struct cpchnl2_event_info *cpchnl2_event = (struct cpchnl2_event_info 
*)msg;
+   struct cpchnl2_vport_info *info;
+   struct cpfl_vport_id vport_identity = { 0 };
+
+   if (msglen < sizeof(struct cpchnl2_event_info)) {
+   PMD_DRV_LOG(ERR, "Error event");
+   return;
+   }
+
+   switch (cpchnl2_event->header.type) {
+   case CPCHNL2_EVENT_VPORT_CREATED:
+   vport_identity.vport_id = 
cpchnl2_event->data.vport_created.vport.vport_id;
+   info = &cpchnl2_event->data.vport_created.info;
+   vport_identity.func_type = info->func_type;
+   vport_identity.pf_id = info->pf_id;
+   vport_identity.vf_id = info->vf_id;
+   if (cpfl_vport_info_create(adapter, &vport_identity, info))
+   PMD_DRV_LOG(WARNING, "Failed to handle 
CPCHNL2_EVENT_VPORT_CREATED");
+   break;
+   case CPCHNL2_EVENT_VPORT_DESTROYED:
+   vport_identity.vport_id = 
cpchnl2_event->data.vport_destroyed.vport.vport_id;
+   vport_identity.func_type = 
cpchnl2_event->data.vport_destroyed.func.func_type;
+   vport_identity.pf_id = 
cpchnl2_event->data.vport_destroyed.func.pf_id;
+   vport_identity.vf_id = 
cpchnl2_event->data.vport_destroyed.func.vf_id;
+   if (cpfl_vport_info_destroy(adapter, &vport_identity))
+   PMD_DRV_LOG(WARNING, "Failed to handle 
CPCHNL2_EVENT_VPORT_DESTROY");
+   break;
+   default:
+   PMD_DRV_LOG(ERR, " unknown event received %u", 
cpchnl2_event->header.type);
+   break;
+   }
+}
+
 static void
 cpfl_handle_virtchnl_msg(struct cpfl_adapter_ext *adapter)
 {
@@ -1535,6 +1638,9 @@ cpfl_handle_virtchnl_msg(struct cpfl_adapter_ext *adapter)
}
cpfl_handle_event_msg(vport, base->mbx_resp,
 

[PATCH v2 05/12] net/cpfl: parse representor devargs

[beilei . xing]

From: Beilei Xing 

Format:

[[c]pf]vf

  controller_id:

  0 : xeon (default)
  1:  acc

  pf_id:

  0 : apf (default)
  1 : cpf

Example:

representor=c0pf0vf[0-3]
  -- xeon > apf > vf 0,1,2,3
 same as pf0vf[0-3] and vf[0-3] if omit default value.

representor=c0pf0
  -- xeon> apf
 same as pf0 if omit default value.

representor=c1pf0
  -- acc > apf

multiple representor devargs are supported.
e.g.: create 4 representors for 4 vfs on xeon APF and one
representor for acc APF.

  -- representor=vf[0-3],representor=c1pf0

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 doc/guides/nics/cpfl.rst   |  36 +
 doc/guides/rel_notes/release_23_11.rst |   3 +
 drivers/net/cpfl/cpfl_ethdev.c | 179 +
 drivers/net/cpfl/cpfl_ethdev.h |   8 ++
 4 files changed, 226 insertions(+)

diff --git a/doc/guides/nics/cpfl.rst b/doc/guides/nics/cpfl.rst
index 39a2b603f3..ff0a183f78 100644
--- a/doc/guides/nics/cpfl.rst
+++ b/doc/guides/nics/cpfl.rst
@@ -92,6 +92,42 @@ Runtime Configuration
   Then the PMD will configure Tx queue with single queue mode.
   Otherwise, split queue mode is chosen by default.
 
+- ``representor`` (default ``not enabled``)
+
+  The cpfl PMD supports the creation of APF/CPF/VF port representors.
+  Each port representor corresponds to a single function of that device.
+  Using the ``devargs`` option ``representor`` the user can specify
+  which functions to create port representors.
+
+  Format is::
+
+[[c]pf]vf
+
+  Controller_id 0 is Xeon (default), while 1 is ACC.
+  Pf_id 0 is APF (default), while 1 is CPF.
+  Default value can be omitted.
+
+  Create 4 representors for 4 vfs on xeon APF::
+
+-a BDF,representor=c0pf0vf[0-3]
+
+  Or::
+
+-a BDF,representor=pf0vf[0-3]
+
+  Or::
+
+-a BDF,representor=vf[0-3]
+
+  Create a representor for CPF on ACC::
+
+-a BDF,representor=c1pf1
+
+  Multiple representor devargs are supported. Create 4 representors for 4
+  vfs on xeon APF and one representor for acc CPF::
+
+-a BDF,representor=vf[0-3],representor=c1pf1
+
 
 Driver compilation and testing
 --
diff --git a/doc/guides/rel_notes/release_23_11.rst 
b/doc/guides/rel_notes/release_23_11.rst
index 4411bb32c1..83826c8896 100644
--- a/doc/guides/rel_notes/release_23_11.rst
+++ b/doc/guides/rel_notes/release_23_11.rst
@@ -72,6 +72,9 @@ New Features
  Also, make sure to start the actual text at the margin.
  ===
 
+* **Updated Intel cpfl driver.**
+
+  * Added support for port representor.
 
 Removed Items
 -
diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 17a69c16fe..a820528a0d 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -13,8 +13,10 @@
 #include 
 
 #include "cpfl_ethdev.h"
+#include 
 #include "cpfl_rxtx.h"
 
+#define CPFL_REPRESENTOR   "representor"
 #define CPFL_TX_SINGLE_Q   "tx_single"
 #define CPFL_RX_SINGLE_Q   "rx_single"
 #define CPFL_VPORT "vport"
@@ -25,6 +27,7 @@ struct cpfl_adapter_list cpfl_adapter_list;
 bool cpfl_adapter_list_init;
 
 static const char * const cpfl_valid_args[] = {
+   CPFL_REPRESENTOR,
CPFL_TX_SINGLE_Q,
CPFL_RX_SINGLE_Q,
CPFL_VPORT,
@@ -1407,6 +1410,128 @@ parse_bool(const char *key, const char *value, void 
*args)
return 0;
 }
 
+static int
+enlist(uint16_t *list, uint16_t *len_list, const uint16_t max_list, uint16_t 
val)
+{
+   uint16_t i;
+
+   for (i = 0; i < *len_list; i++) {
+   if (list[i] == val)
+   return 0;
+   }
+   if (*len_list >= max_list)
+   return -1;
+   list[(*len_list)++] = val;
+   return 0;
+}
+
+static const char *
+process_range(const char *str, uint16_t *list, uint16_t *len_list,
+   const uint16_t max_list)
+{
+   uint16_t lo, hi, val;
+   int result, n = 0;
+   const char *pos = str;
+
+   result = sscanf(str, "%hu%n-%hu%n", &lo, &n, &hi, &n);
+   if (result == 1) {
+   if (enlist(list, len_list, max_list, lo) != 0)
+   return NULL;
+   } else if (result == 2) {
+   if (lo > hi)
+   return NULL;
+   for (val = lo; val <= hi; val++) {
+   if (enlist(list, len_list, max_list, val) != 0)
+   return NULL;
+   }
+   } else {
+   return NULL;
+   }
+   return pos + n;
+}
+
+static const char *
+process_list(const char *str, uint16_t *list, uint16_t *len_list, const 
uint16_t max_list)
+{
+   const char *pos = str;
+
+   if (*pos == '[')
+   pos++;
+   while (1) {
+   pos = process_range(pos, list, len_list, max_list);
+   if (pos == NULL)
+   return NULL;
+   if (*pos != ',') /* end of list */
+  

[PATCH v2 06/12] net/cpfl: support probe again

[beilei . xing]

From: Beilei Xing 

Only representor will be parsed for probe again.

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c | 69 +++---
 1 file changed, 56 insertions(+), 13 deletions(-)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index a820528a0d..09015fbb08 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -26,7 +26,7 @@ rte_spinlock_t cpfl_adapter_lock;
 struct cpfl_adapter_list cpfl_adapter_list;
 bool cpfl_adapter_list_init;
 
-static const char * const cpfl_valid_args[] = {
+static const char * const cpfl_valid_args_first[] = {
CPFL_REPRESENTOR,
CPFL_TX_SINGLE_Q,
CPFL_RX_SINGLE_Q,
@@ -34,6 +34,11 @@ static const char * const cpfl_valid_args[] = {
NULL
 };
 
+static const char * const cpfl_valid_args_again[] = {
+   CPFL_REPRESENTOR,
+   NULL
+};
+
 uint32_t cpfl_supported_speeds[] = {
RTE_ETH_SPEED_NUM_NONE,
RTE_ETH_SPEED_NUM_10M,
@@ -1533,7 +1538,7 @@ parse_repr(const char *key __rte_unused, const char 
*value, void *args)
 }
 
 static int
-cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
+cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter, bool first)
 {
struct rte_devargs *devargs = pci_dev->device.devargs;
struct cpfl_devargs *cpfl_args = &adapter->devargs;
@@ -1545,7 +1550,8 @@ cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct 
cpfl_adapter_ext *adap
if (devargs == NULL)
return 0;
 
-   kvlist = rte_kvargs_parse(devargs->args, cpfl_valid_args);
+   kvlist = rte_kvargs_parse(devargs->args,
+   first ? cpfl_valid_args_first : cpfl_valid_args_again);
if (kvlist == NULL) {
PMD_INIT_LOG(ERR, "invalid kvargs key");
return -EINVAL;
@@ -1562,6 +1568,9 @@ cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct 
cpfl_adapter_ext *adap
if (ret != 0)
goto fail;
 
+   if (!first)
+   return 0;
+
ret = rte_kvargs_process(kvlist, CPFL_VPORT, &parse_vport,
 cpfl_args);
if (ret != 0)
@@ -2291,18 +2300,11 @@ cpfl_vport_create(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *adapt
 }
 
 static int
-cpfl_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
-  struct rte_pci_device *pci_dev)
+cpfl_pci_probe_first(struct rte_pci_device *pci_dev)
 {
struct cpfl_adapter_ext *adapter;
int retval;
 
-   if (!cpfl_adapter_list_init) {
-   rte_spinlock_init(&cpfl_adapter_lock);
-   TAILQ_INIT(&cpfl_adapter_list);
-   cpfl_adapter_list_init = true;
-   }
-
adapter = rte_zmalloc("cpfl_adapter_ext",
  sizeof(struct cpfl_adapter_ext), 0);
if (adapter == NULL) {
@@ -2310,7 +2312,7 @@ cpfl_pci_probe(struct rte_pci_driver *pci_drv 
__rte_unused,
return -ENOMEM;
}
 
-   retval = cpfl_parse_devargs(pci_dev, adapter);
+   retval = cpfl_parse_devargs(pci_dev, adapter, true);
if (retval != 0) {
PMD_INIT_LOG(ERR, "Failed to parse private devargs");
return retval;
@@ -2355,6 +2357,46 @@ cpfl_pci_probe(struct rte_pci_driver *pci_drv 
__rte_unused,
return retval;
 }
 
+static int
+cpfl_pci_probe_again(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
+{
+   int ret;
+
+   ret = cpfl_parse_devargs(pci_dev, adapter, false);
+   if (ret != 0) {
+   PMD_INIT_LOG(ERR, "Failed to parse private devargs");
+   return ret;
+   }
+
+   ret = cpfl_repr_devargs_process(adapter);
+   if (ret != 0) {
+   PMD_INIT_LOG(ERR, "Failed to process reprenstor devargs");
+   return ret;
+   }
+
+   return 0;
+}
+
+static int
+cpfl_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
+  struct rte_pci_device *pci_dev)
+{
+   struct cpfl_adapter_ext *adapter;
+
+   if (!cpfl_adapter_list_init) {
+   rte_spinlock_init(&cpfl_adapter_lock);
+   TAILQ_INIT(&cpfl_adapter_list);
+   cpfl_adapter_list_init = true;
+   }
+
+   adapter = cpfl_find_adapter_ext(pci_dev);
+
+   if (adapter == NULL)
+   return cpfl_pci_probe_first(pci_dev);
+   else
+   return cpfl_pci_probe_again(pci_dev, adapter);
+}
+
 static int
 cpfl_pci_remove(struct rte_pci_device *pci_dev)
 {
@@ -2377,7 +2419,8 @@ cpfl_pci_remove(struct rte_pci_device *pci_dev)
 
 static struct rte_pci_driver rte_cpfl_pmd = {
.id_table   = pci_id_cpfl_map,
-   .drv_flags  = RTE_PCI_DRV_NEED_MAPPING,
+   .drv_flags  = RTE_PCI_DRV_NEED_MAPPING |
+ RTE_PCI_DRV_PROBE_AGAIN,
.probe  = cpfl_pci_probe,
  

[PATCH v2 07/12] net/cpfl: create port representor

[beilei . xing]

From: Beilei Xing 

Track representor request in a whitelist.
Representor will only be created for active vport.

Signed-off-by: Jingjing Wu 
Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c  | 107 ---
 drivers/net/cpfl/cpfl_ethdev.h  |  34 +++
 drivers/net/cpfl/cpfl_representor.c | 448 
 drivers/net/cpfl/cpfl_representor.h |  26 ++
 drivers/net/cpfl/meson.build|   1 +
 5 files changed, 573 insertions(+), 43 deletions(-)
 create mode 100644 drivers/net/cpfl/cpfl_representor.c
 create mode 100644 drivers/net/cpfl/cpfl_representor.h

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 09015fbb08..08daade7ac 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1898,6 +1898,42 @@ cpfl_vport_map_uninit(struct cpfl_adapter_ext *adapter)
rte_hash_free(adapter->vport_map_hash);
 }
 
+static int
+cpfl_repr_allowlist_init(struct cpfl_adapter_ext *adapter)
+{
+   char hname[32];
+
+   snprintf(hname, 32, "%s-repr_wl", adapter->name);
+
+   rte_spinlock_init(&adapter->repr_lock);
+
+#define CPFL_REPR_HASH_ENTRY_NUM 2048
+
+   struct rte_hash_parameters params = {
+   .name = hname,
+   .entries = CPFL_REPR_HASH_ENTRY_NUM,
+   .key_len = sizeof(struct cpfl_repr_id),
+   .hash_func = rte_hash_crc,
+   .socket_id = SOCKET_ID_ANY,
+   };
+
+   adapter->repr_allowlist_hash = rte_hash_create(¶ms);
+
+   if (adapter->repr_allowlist_hash == NULL) {
+   PMD_INIT_LOG(ERR, "Failed to create repr allowlist hash");
+   return -EINVAL;
+   }
+
+   return 0;
+}
+
+static void
+cpfl_repr_allowlist_uninit(struct cpfl_adapter_ext *adapter)
+{
+   rte_hash_free(adapter->repr_allowlist_hash);
+}
+
+
 static int
 cpfl_adapter_ext_init(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
 {
@@ -1928,6 +1964,12 @@ cpfl_adapter_ext_init(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *a
goto err_vport_map_init;
}
 
+   ret = cpfl_repr_allowlist_init(adapter);
+   if (ret) {
+   PMD_INIT_LOG(ERR, "Failed to init representor allowlist");
+   goto err_repr_allowlist_init;
+   }
+
rte_eal_alarm_set(CPFL_ALARM_INTERVAL, cpfl_dev_alarm_handler, adapter);
 
adapter->max_vport_nb = adapter->base.caps.max_vports > 
CPFL_MAX_VPORT_NUM ?
@@ -1952,6 +1994,8 @@ cpfl_adapter_ext_init(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *a
 
 err_vports_alloc:
rte_eal_alarm_cancel(cpfl_dev_alarm_handler, adapter);
+   cpfl_repr_allowlist_uninit(adapter);
+err_repr_allowlist_init:
cpfl_vport_map_uninit(adapter);
 err_vport_map_init:
idpf_adapter_deinit(base);
@@ -2227,48 +2271,6 @@ cpfl_vport_devargs_process(struct cpfl_adapter_ext 
*adapter)
return 0;
 }
 
-static int
-cpfl_repr_devargs_process(struct cpfl_adapter_ext *adapter)
-{
-   struct cpfl_devargs *devargs = &adapter->devargs;
-   int i, j;
-
-   /* check and refine repr args */
-   for (i = 0; i < devargs->repr_args_num; i++) {
-   struct rte_eth_devargs *eth_da = &devargs->repr_args[i];
-
-   /* set default host_id to xeon host */
-   if (eth_da->nb_mh_controllers == 0) {
-   eth_da->nb_mh_controllers = 1;
-   eth_da->mh_controllers[0] = CPFL_HOST_ID_HOST;
-   } else {
-   for (j = 0; j < eth_da->nb_mh_controllers; j++) {
-   if (eth_da->mh_controllers[j] > 
CPFL_HOST_ID_ACC) {
-   PMD_INIT_LOG(ERR, "Invalid Host ID %d",
-eth_da->mh_controllers[j]);
-   return -EINVAL;
-   }
-   }
-   }
-
-   /* set default pf to APF */
-   if (eth_da->nb_ports == 0) {
-   eth_da->nb_ports = 1;
-   eth_da->ports[0] = CPFL_PF_TYPE_APF;
-   } else {
-   for (j = 0; j < eth_da->nb_ports; j++) {
-   if (eth_da->ports[j] > CPFL_PF_TYPE_CPF) {
-   PMD_INIT_LOG(ERR, "Invalid Host ID %d",
-eth_da->ports[j]);
-   return -EINVAL;
-   }
-   }
-   }
-   }
-
-   return 0;
-}
-
 static int
 cpfl_vport_create(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
 {
@@ -2304,6 +2306,7 @@ cpfl_pci_probe_first(struct rte_pci_device *pci_dev)
 {
struct cpfl_adapter_ext *adapter;
int retval;
+   uint16_t port_id;
 
adapter = rte_zmalloc("cp

[PATCH v2 08/12] net/cpfl: support vport list/info get

[beilei . xing]

From: Beilei Xing 

Support cp channel ops CPCHNL2_OP_CPF_GET_VPORT_LIST and
CPCHNL2_OP_CPF_GET_VPORT_INFO.

Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.h |  8 
 drivers/net/cpfl/cpfl_vchnl.c  | 72 ++
 drivers/net/cpfl/meson.build   |  1 +
 3 files changed, 81 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_vchnl.c

diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index d4d9727a80..4f6944d00a 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -189,6 +189,14 @@ struct cpfl_adapter_ext {
 
 TAILQ_HEAD(cpfl_adapter_list, cpfl_adapter_ext);
 
+int cpfl_cc_vport_list_get(struct cpfl_adapter_ext *adapter,
+  struct cpfl_vport_id *vi,
+  struct cpchnl2_get_vport_list_response *response);
+int cpfl_cc_vport_info_get(struct cpfl_adapter_ext *adapter,
+  struct cpchnl2_vport_id *vport_id,
+  struct cpfl_vport_id *vi,
+  struct cpchnl2_get_vport_info_response *response);
+
 #define CPFL_DEV_TO_PCI(eth_dev)   \
RTE_DEV_TO_PCI((eth_dev)->device)
 #define CPFL_ADAPTER_TO_EXT(p) \
diff --git a/drivers/net/cpfl/cpfl_vchnl.c b/drivers/net/cpfl/cpfl_vchnl.c
new file mode 100644
index 00..a21a4a451f
--- /dev/null
+++ b/drivers/net/cpfl/cpfl_vchnl.c
@@ -0,0 +1,72 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#include "cpfl_ethdev.h"
+#include 
+
+int
+cpfl_cc_vport_list_get(struct cpfl_adapter_ext *adapter,
+  struct cpfl_vport_id *vi,
+  struct cpchnl2_get_vport_list_response *response)
+{
+   struct cpchnl2_get_vport_list_request request;
+   struct idpf_cmd_info args;
+   int err;
+
+   memset(&request, 0, sizeof(request));
+   request.func_type = vi->func_type;
+   request.pf_id = vi->pf_id;
+   request.vf_id = vi->vf_id;
+
+   memset(&args, 0, sizeof(args));
+   args.ops = CPCHNL2_OP_GET_VPORT_LIST;
+   args.in_args = (uint8_t *)&request;
+   args.in_args_size = sizeof(struct cpchnl2_get_vport_list_request);
+   args.out_buffer = adapter->base.mbx_resp;
+   args.out_size = IDPF_DFLT_MBX_BUF_SIZE;
+
+   err = idpf_vc_cmd_execute(&adapter->base, &args);
+   if (err != 0) {
+   PMD_DRV_LOG(ERR, "Failed to execute command of 
CPCHNL2_OP_GET_VPORT_LIST");
+   return err;
+   }
+
+   rte_memcpy(response, args.out_buffer, IDPF_DFLT_MBX_BUF_SIZE);
+
+   return 0;
+}
+
+int
+cpfl_cc_vport_info_get(struct cpfl_adapter_ext *adapter,
+  struct cpchnl2_vport_id *vport_id,
+  struct cpfl_vport_id *vi,
+  struct cpchnl2_get_vport_info_response *response)
+{
+   struct cpchnl2_get_vport_info_request request;
+   struct idpf_cmd_info args;
+   int err;
+
+   request.vport.vport_id = vport_id->vport_id;
+   request.vport.vport_type = vport_id->vport_type;
+   request.func.func_type = vi->func_type;
+   request.func.pf_id = vi->pf_id;
+   request.func.vf_id = vi->vf_id;
+
+   memset(&args, 0, sizeof(args));
+   args.ops = CPCHNL2_OP_GET_VPORT_INFO;
+   args.in_args = (uint8_t *)&request;
+   args.in_args_size = sizeof(struct cpchnl2_get_vport_info_request);
+   args.out_buffer = adapter->base.mbx_resp;
+   args.out_size = IDPF_DFLT_MBX_BUF_SIZE;
+
+   err = idpf_vc_cmd_execute(&adapter->base, &args);
+   if (err != 0) {
+   PMD_DRV_LOG(ERR, "Failed to execute command of 
CPCHNL2_OP_GET_VPORT_INFO");
+   return err;
+   }
+
+   rte_memcpy(response, args.out_buffer, sizeof(*response));
+
+   return 0;
+}
diff --git a/drivers/net/cpfl/meson.build b/drivers/net/cpfl/meson.build
index 1d963e5fd1..fb075c6860 100644
--- a/drivers/net/cpfl/meson.build
+++ b/drivers/net/cpfl/meson.build
@@ -17,6 +17,7 @@ sources = files(
 'cpfl_ethdev.c',
 'cpfl_rxtx.c',
 'cpfl_representor.c',
+'cpfl_vchnl.c',
 )
 
 if arch_subdir == 'x86'
-- 
2.34.1

[PATCH v2 09/12] net/cpfl: update vport info before creating representor

[beilei . xing]

From: Beilei Xing 

Get port representor's vport list and update vport_map_hash
before creating the port representor.

Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c  |   2 +-
 drivers/net/cpfl/cpfl_ethdev.h  |   3 +
 drivers/net/cpfl/cpfl_representor.c | 124 
 3 files changed, 128 insertions(+), 1 deletion(-)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 08daade7ac..e552387cfe 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1633,7 +1633,7 @@ cpfl_handle_event_msg(struct idpf_vport *vport, uint8_t 
*msg, uint16_t msglen)
}
 }
 
-static int
+int
 cpfl_vport_info_create(struct cpfl_adapter_ext *adapter,
   struct cpfl_vport_id *vport_identity,
   struct cpchnl2_vport_info *vport_info)
diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index 4f6944d00a..cc7f43fc3e 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -189,6 +189,9 @@ struct cpfl_adapter_ext {
 
 TAILQ_HEAD(cpfl_adapter_list, cpfl_adapter_ext);
 
+int cpfl_vport_info_create(struct cpfl_adapter_ext *adapter,
+  struct cpfl_vport_id *vport_identity,
+  struct cpchnl2_vport_info *vport_info);
 int cpfl_cc_vport_list_get(struct cpfl_adapter_ext *adapter,
   struct cpfl_vport_id *vi,
   struct cpchnl2_get_vport_list_response *response);
diff --git a/drivers/net/cpfl/cpfl_representor.c 
b/drivers/net/cpfl/cpfl_representor.c
index 29da56dda0..ed2d1fff17 100644
--- a/drivers/net/cpfl/cpfl_representor.c
+++ b/drivers/net/cpfl/cpfl_representor.c
@@ -368,6 +368,86 @@ match_repr_with_vport(const struct cpfl_repr_id *repr_id,
return false;
 }
 
+static int
+cpfl_repr_vport_list_query(struct cpfl_adapter_ext *adapter,
+  const struct cpfl_repr_id *repr_id,
+  struct cpchnl2_get_vport_list_response *response)
+{
+   struct cpfl_vport_id vi;
+   int ret;
+
+   if (repr_id->type == RTE_ETH_REPRESENTOR_PF) {
+   /* PF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_PF;
+   vi.pf_id = cpfl_func_id_get(repr_id->host_id, repr_id->pf_id);
+   vi.vf_id = 0;
+   } else {
+   /* VF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_SRIOV;
+   vi.pf_id = HOST0_APF;
+   vi.vf_id = repr_id->vf_id;
+   }
+
+   ret = cpfl_cc_vport_list_get(adapter, &vi, response);
+
+   return ret;
+}
+
+static int
+cpfl_repr_vport_info_query(struct cpfl_adapter_ext *adapter,
+  const struct cpfl_repr_id *repr_id,
+  struct cpchnl2_vport_id *vport_id,
+  struct cpchnl2_get_vport_info_response *response)
+{
+   struct cpfl_vport_id vi;
+   int ret;
+
+   if (repr_id->type == RTE_ETH_REPRESENTOR_PF) {
+   /* PF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_PF;
+   vi.pf_id = cpfl_func_id_get(repr_id->host_id, repr_id->pf_id);
+   vi.vf_id = 0;
+   } else {
+   /* VF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_SRIOV;
+   vi.pf_id = HOST0_APF;
+   vi.vf_id = repr_id->vf_id;
+   }
+
+   ret = cpfl_cc_vport_info_get(adapter, vport_id, &vi, response);
+
+   return ret;
+}
+
+static int
+cpfl_repr_vport_map_update(struct cpfl_adapter_ext *adapter,
+  const struct cpfl_repr_id *repr_id, uint32_t 
vport_id,
+  struct cpchnl2_get_vport_info_response *response)
+{
+   struct cpfl_vport_id vi;
+   int ret;
+
+   vi.vport_id = vport_id;
+   if (repr_id->type == RTE_ETH_REPRESENTOR_PF) {
+   /* PF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_PF;
+   vi.pf_id = cpfl_func_id_get(repr_id->host_id, repr_id->pf_id);
+   } else {
+   /* VF */
+   vi.func_type = CPCHNL2_FUNC_TYPE_SRIOV;
+   vi.pf_id = HOST0_APF;
+   vi.vf_id = repr_id->vf_id;
+   }
+
+   ret = cpfl_vport_info_create(adapter, &vi, &response->info);
+   if (ret != 0) {
+   PMD_INIT_LOG(ERR, "Fail to update vport map hash for 
representor.");
+   return ret;
+   }
+
+   return 0;
+}
+
 int
 cpfl_repr_create(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter)
 {
@@ -375,8 +455,14 @@ cpfl_repr_create(struct rte_pci_device *pci_dev, struct 
cpfl_adapter_ext *adapte
uint32_t iter = 0;
const struct cpfl_repr_id *repr_id;
const struct cpfl_vport_id *vp_id;
+   struct cpchnl2_get_vport_list_response *vlist_resp;
+   struct cpchnl2_get_vport_info_response vinfo_resp;
int ret;
 
+   vlist_resp = rte_zmalloc(NULL, IDPF_DFLT_MBX_BUF_SIZE, 0);
+   

[PATCH v2 10/12] net/cpfl: refine handle virtual channel message

[beilei . xing]

From: Beilei Xing 

Refine handle virtual channel event message.

Signed-off-by: Qi Zhang 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.c | 46 --
 1 file changed, 22 insertions(+), 24 deletions(-)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index e552387cfe..330a865e3c 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1591,40 +1591,50 @@ cpfl_parse_devargs(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *adap
return ret;
 }
 
-static struct idpf_vport *
+static struct cpfl_vport *
 cpfl_find_vport(struct cpfl_adapter_ext *adapter, uint32_t vport_id)
 {
-   struct idpf_vport *vport = NULL;
+   struct cpfl_vport *vport = NULL;
int i;
 
for (i = 0; i < adapter->cur_vport_nb; i++) {
-   vport = &adapter->vports[i]->base;
-   if (vport->vport_id != vport_id)
+   vport = adapter->vports[i];
+   if (vport->base.vport_id != vport_id)
continue;
else
return vport;
}
 
-   return vport;
+   return NULL;
 }
 
 static void
-cpfl_handle_event_msg(struct idpf_vport *vport, uint8_t *msg, uint16_t msglen)
+cpfl_handle_vchnl_event_msg(struct cpfl_adapter_ext *adapter, uint8_t *msg, 
uint16_t msglen)
 {
struct virtchnl2_event *vc_event = (struct virtchnl2_event *)msg;
-   struct rte_eth_dev_data *data = vport->dev_data;
-   struct rte_eth_dev *dev = &rte_eth_devices[data->port_id];
+   struct cpfl_vport *vport;
+   struct rte_eth_dev_data *data;
+   struct rte_eth_dev *dev;
 
if (msglen < sizeof(struct virtchnl2_event)) {
PMD_DRV_LOG(ERR, "Error event");
return;
}
 
+   vport = cpfl_find_vport(adapter, vc_event->vport_id);
+   if (!vport) {
+   PMD_DRV_LOG(ERR, "Can't find vport.");
+   return;
+   }
+
+   data = vport->itf.data;
+   dev = &rte_eth_devices[data->port_id];
+
switch (vc_event->event) {
case VIRTCHNL2_EVENT_LINK_CHANGE:
PMD_DRV_LOG(DEBUG, "VIRTCHNL2_EVENT_LINK_CHANGE");
-   vport->link_up = !!(vc_event->link_status);
-   vport->link_speed = vc_event->link_speed;
+   vport->base.link_up = !!(vc_event->link_status);
+   vport->base.link_speed = vc_event->link_speed;
cpfl_dev_link_update(dev, 0);
break;
default:
@@ -1741,10 +1751,8 @@ cpfl_handle_virtchnl_msg(struct cpfl_adapter_ext 
*adapter)
struct idpf_adapter *base = &adapter->base;
struct idpf_dma_mem *dma_mem = NULL;
struct idpf_hw *hw = &base->hw;
-   struct virtchnl2_event *vc_event;
struct idpf_ctlq_msg ctlq_msg;
enum idpf_mbx_opc mbx_op;
-   struct idpf_vport *vport;
uint16_t pending = 1;
uint32_t vc_op;
int ret;
@@ -1766,18 +1774,8 @@ cpfl_handle_virtchnl_msg(struct cpfl_adapter_ext 
*adapter)
switch (mbx_op) {
case idpf_mbq_opc_send_msg_to_peer_pf:
if (vc_op == VIRTCHNL2_OP_EVENT) {
-   if (ctlq_msg.data_len < sizeof(struct 
virtchnl2_event)) {
-   PMD_DRV_LOG(ERR, "Error event");
-   return;
-   }
-   vc_event = (struct virtchnl2_event 
*)base->mbx_resp;
-   vport = cpfl_find_vport(adapter, 
vc_event->vport_id);
-   if (!vport) {
-   PMD_DRV_LOG(ERR, "Can't find vport.");
-   return;
-   }
-   cpfl_handle_event_msg(vport, base->mbx_resp,
- ctlq_msg.data_len);
+   cpfl_handle_vchnl_event_msg(adapter, 
adapter->base.mbx_resp,
+   ctlq_msg.data_len);
} else if (vc_op == CPCHNL2_OP_EVENT) {
cpfl_handle_cpchnl_event_msg(adapter, 
adapter->base.mbx_resp,
 ctlq_msg.data_len);
-- 
2.34.1

[PATCH v2 11/12] net/cpfl: support link update for representor

[beilei . xing]

From: Beilei Xing 

Add link update ops for representor.

Signed-off-by: Jingjing Wu 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_ethdev.h  |  1 +
 drivers/net/cpfl/cpfl_representor.c | 21 +
 2 files changed, 22 insertions(+)

diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index cc7f43fc3e..55bd119423 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -163,6 +163,7 @@ struct cpfl_repr {
struct cpfl_repr_id repr_id;
struct rte_ether_addr mac_addr;
struct cpfl_vport_info *vport_info;
+   bool func_up; /* If the represented function is up */
 };
 
 struct cpfl_adapter_ext {
diff --git a/drivers/net/cpfl/cpfl_representor.c 
b/drivers/net/cpfl/cpfl_representor.c
index ed2d1fff17..5b5c959727 100644
--- a/drivers/net/cpfl/cpfl_representor.c
+++ b/drivers/net/cpfl/cpfl_representor.c
@@ -285,12 +285,31 @@ cpfl_repr_dev_stop(struct rte_eth_dev *dev)
return 0;
 }
 
+static int
+cpfl_repr_link_update(struct rte_eth_dev *ethdev,
+ __rte_unused int wait_to_complete)
+{
+   struct cpfl_repr *repr = CPFL_DEV_TO_REPR(ethdev);
+   struct rte_eth_link *dev_link = ðdev->data->dev_link;
+
+   if (!(ethdev->data->dev_flags & RTE_ETH_DEV_REPRESENTOR)) {
+   PMD_INIT_LOG(ERR, "This ethdev is not representor.");
+   return -EINVAL;
+   }
+   dev_link->link_status = repr->func_up ?
+   RTE_ETH_LINK_UP : RTE_ETH_LINK_DOWN;
+
+   return 0;
+}
+
 static const struct eth_dev_ops cpfl_repr_dev_ops = {
.dev_start  = cpfl_repr_dev_start,
.dev_stop   = cpfl_repr_dev_stop,
.dev_configure  = cpfl_repr_dev_configure,
.dev_close  = cpfl_repr_dev_close,
.dev_infos_get  = cpfl_repr_dev_info_get,
+
+   .link_update= cpfl_repr_link_update,
 };
 
 static int
@@ -305,6 +324,8 @@ cpfl_repr_init(struct rte_eth_dev *eth_dev, void 
*init_param)
repr->itf.type = CPFL_ITF_TYPE_REPRESENTOR;
repr->itf.adapter = adapter;
repr->itf.data = eth_dev->data;
+   if (repr->vport_info->vport_info.vport_status == 
CPCHNL2_VPORT_STATUS_ENABLED)
+   repr->func_up = true;
 
eth_dev->dev_ops = &cpfl_repr_dev_ops;
 
-- 
2.34.1

[PATCH v2 12/12] net/cpfl: support Rx/Tx queue setup for representor

[beilei . xing]

From: Beilei Xing 

Add dummy Rx/Tx queue setup functions for representor.

Signed-off-by: Jingjing Wu 
Signed-off-by: Beilei Xing 
---
 drivers/net/cpfl/cpfl_representor.c | 26 ++
 1 file changed, 26 insertions(+)

diff --git a/drivers/net/cpfl/cpfl_representor.c 
b/drivers/net/cpfl/cpfl_representor.c
index 5b5c959727..58e0d91d97 100644
--- a/drivers/net/cpfl/cpfl_representor.c
+++ b/drivers/net/cpfl/cpfl_representor.c
@@ -285,6 +285,29 @@ cpfl_repr_dev_stop(struct rte_eth_dev *dev)
return 0;
 }
 
+static int
+idpf_repr_rx_queue_setup(__rte_unused struct rte_eth_dev *dev,
+__rte_unused uint16_t queue_id,
+__rte_unused uint16_t nb_desc,
+__rte_unused unsigned int socket_id,
+__rte_unused const struct rte_eth_rxconf *conf,
+__rte_unused struct rte_mempool *pool)
+{
+   /* Dummy */
+   return 0;
+}
+
+static int
+idpf_repr_tx_queue_setup(__rte_unused struct rte_eth_dev *dev,
+__rte_unused uint16_t queue_id,
+__rte_unused uint16_t nb_desc,
+__rte_unused unsigned int socket_id,
+__rte_unused const struct rte_eth_txconf *conf)
+{
+   /* Dummy */
+   return 0;
+}
+
 static int
 cpfl_repr_link_update(struct rte_eth_dev *ethdev,
  __rte_unused int wait_to_complete)
@@ -309,6 +332,9 @@ static const struct eth_dev_ops cpfl_repr_dev_ops = {
.dev_close  = cpfl_repr_dev_close,
.dev_infos_get  = cpfl_repr_dev_info_get,
 
+   .rx_queue_setup = idpf_repr_rx_queue_setup,
+   .tx_queue_setup = idpf_repr_tx_queue_setup,
+
.link_update= cpfl_repr_link_update,
 };
 
-- 
2.34.1