[RFC,v2,04/10] dts: add dpdk execution handling

Add methods for setting up and shutting down DPDK apps and for
constructing EAL parameters.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
---
 dts/conf.yaml                                 |   4 +
 dts/framework/config/__init__.py              |  85 ++++++++-
 dts/framework/config/conf_yaml_schema.json    |  22 +++
 .../remote_session/os/linux_session.py        |  15 ++
 dts/framework/remote_session/os/os_session.py |  16 +-
 .../remote_session/os/posix_session.py        |  80 ++++++++
 dts/framework/testbed_model/hw/__init__.py    |  17 ++
 dts/framework/testbed_model/hw/cpu.py         | 164 ++++++++++++++++
 dts/framework/testbed_model/node/node.py      |  36 ++++
 dts/framework/testbed_model/node/sut_node.py  | 178 +++++++++++++++++-
 dts/framework/utils.py                        |  20 ++
 11 files changed, 634 insertions(+), 3 deletions(-)
 create mode 100644 dts/framework/testbed_model/hw/__init__.py
 create mode 100644 dts/framework/testbed_model/hw/cpu.py
  

On Mon, Nov 14, 2022 at 11:54 AM Juraj Linkeš <juraj.linkes@pantheon.tech>
wrote:

> Add methods for setting up and shutting down DPDK apps and for
> constructing EAL parameters.
>
> Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
> ---
>  dts/conf.yaml                                 |   4 +
>  dts/framework/config/__init__.py              |  85 ++++++++-
>  dts/framework/config/conf_yaml_schema.json    |  22 +++
>  .../remote_session/os/linux_session.py        |  15 ++
>  dts/framework/remote_session/os/os_session.py |  16 +-
>  .../remote_session/os/posix_session.py        |  80 ++++++++
>  dts/framework/testbed_model/hw/__init__.py    |  17 ++
>  dts/framework/testbed_model/hw/cpu.py         | 164 ++++++++++++++++
>  dts/framework/testbed_model/node/node.py      |  36 ++++
>  dts/framework/testbed_model/node/sut_node.py  | 178 +++++++++++++++++-
>  dts/framework/utils.py                        |  20 ++
>  11 files changed, 634 insertions(+), 3 deletions(-)
>  create mode 100644 dts/framework/testbed_model/hw/__init__.py
>  create mode 100644 dts/framework/testbed_model/hw/cpu.py
>
> diff --git a/dts/conf.yaml b/dts/conf.yaml
> index 6b0bc5c2bf..976888a88e 100644
> --- a/dts/conf.yaml
> +++ b/dts/conf.yaml
> @@ -12,4 +12,8 @@ nodes:
>    - name: "SUT 1"
>      hostname: sut1.change.me.localhost
>      user: root
> +    arch: x86_64
>      os: linux
> +    bypass_core0: true
> +    cpus: ""
> +    memory_channels: 4
> diff --git a/dts/framework/config/__init__.py
> b/dts/framework/config/__init__.py
> index 1b97dc3ab9..344d697a69 100644
> --- a/dts/framework/config/__init__.py
> +++ b/dts/framework/config/__init__.py
> @@ -11,12 +11,13 @@
>  import pathlib
>  from dataclasses import dataclass
>  from enum import Enum, auto, unique
> -from typing import Any
> +from typing import Any, Iterable
>
>  import warlock  # type: ignore
>  import yaml
>
>  from framework.settings import SETTINGS
> +from framework.utils import expand_range
>
>
>  class StrEnum(Enum):
> @@ -60,6 +61,80 @@ class Compiler(StrEnum):
>      msvc = auto()
>
>
> +@dataclass(slots=True, frozen=True)
> +class CPU:
> +    cpu: int
> +    core: int
> +    socket: int
> +    node: int
> +
> +    def __str__(self) -> str:
> +        return str(self.cpu)
> +
> +
> +class CPUList(object):
> +    """
> +    Convert these options into a list of int cpus
> +    cpu_list=[CPU1, CPU2] - a list of CPUs
> +    cpu_list=[0,1,2,3] - a list of int indices
> +    cpu_list=['0','1','2-3'] - a list of str indices; ranges are supported
> +    cpu_list='0,1,2-3' - a comma delimited str of indices; ranges are
> supported
> +
> +    The class creates a unified format used across the framework and
> allows
> +    the user to use either a str representation (using str(instance) or
> directly
> +    in f-strings) or a list representation (by accessing
> instance.cpu_list).
> +    Empty cpu_list is allowed.
> +    """
> +
> +    _cpu_list: list[int]
> +
> +    def __init__(self, cpu_list: list[int | str | CPU] | str):
> +        self._cpu_list = []
> +        if isinstance(cpu_list, str):
> +            self._from_str(cpu_list.split(","))
> +        else:
> +            self._from_str((str(cpu) for cpu in cpu_list))
> +
> +        # the input cpus may not be sorted
> +        self._cpu_list.sort()
> +
> +    @property
> +    def cpu_list(self) -> list[int]:
> +        return self._cpu_list
> +
> +    def _from_str(self, cpu_list: Iterable[str]) -> None:
> +        for cpu in cpu_list:
> +            self._cpu_list.extend(expand_range(cpu))
> +
> +    def _get_consecutive_cpus_range(self, cpu_list: list[int]) ->
> list[str]:
> +        formatted_core_list = []
> +        tmp_cpus_list = list(sorted(cpu_list))
> +        segment = tmp_cpus_list[:1]
> +        for core_id in tmp_cpus_list[1:]:
> +            if core_id - segment[-1] == 1:
> +                segment.append(core_id)
> +            else:
> +                formatted_core_list.append(
> +                    f"{segment[0]}-{segment[-1]}"
> +                    if len(segment) > 1
> +                    else f"{segment[0]}"
> +                )
> +                current_core_index = tmp_cpus_list.index(core_id)
> +                formatted_core_list.extend(
> +
> self._get_consecutive_cpus_range(tmp_cpus_list[current_core_index:])
> +                )
> +                segment.clear()
> +                break
> +        if len(segment) > 0:
> +            formatted_core_list.append(
> +                f"{segment[0]}-{segment[-1]}" if len(segment) > 1 else
> f"{segment[0]}"
> +            )
> +        return formatted_core_list
> +
> +    def __str__(self) -> str:
> +        return
> f'{",".join(self._get_consecutive_cpus_range(self._cpu_list))}'
> +
> +
>  # Slots enables some optimizations, by pre-allocating space for the
> defined
>  # attributes in the underlying data structure.
>  #
> @@ -71,7 +146,11 @@ class NodeConfiguration:
>      hostname: str
>      user: str
>      password: str | None
> +    arch: Architecture
>      os: OS
> +    bypass_core0: bool
> +    cpus: CPUList
> +    memory_channels: int
>
>      @staticmethod
>      def from_dict(d: dict) -> "NodeConfiguration":
> @@ -80,7 +159,11 @@ def from_dict(d: dict) -> "NodeConfiguration":
>              hostname=d["hostname"],
>              user=d["user"],
>              password=d.get("password"),
> +            arch=Architecture(d["arch"]),
>              os=OS(d["os"]),
> +            bypass_core0=d.get("bypass_core0", False),
> +            cpus=CPUList(d.get("cpus", "1")),
> +            memory_channels=d.get("memory_channels", 1),
>          )
>
>
> diff --git a/dts/framework/config/conf_yaml_schema.json
> b/dts/framework/config/conf_yaml_schema.json
> index 409ce7ac74..c59d3e30e6 100644
> --- a/dts/framework/config/conf_yaml_schema.json
> +++ b/dts/framework/config/conf_yaml_schema.json
> @@ -6,6 +6,12 @@
>        "type": "string",
>        "description": "A unique identifier for a node"
>      },
> +    "ARCH": {
> +      "type": "string",
> +      "enum": [
> +        "x86_64"
>

arm64 and ppc64le should probably be included here. I think that we can
focus on 64 bit arches for now.


> +      ]
> +    },
>      "OS": {
>        "type": "string",
>        "enum": [
> @@ -82,8 +88,23 @@
>              "type": "string",
>              "description": "The password to use on this node. Use only as
> a last resort. SSH keys are STRONGLY preferred."
>            },
> +          "arch": {
> +            "$ref": "#/definitions/ARCH"
> +          },
>            "os": {
>              "$ref": "#/definitions/OS"
> +          },
> +          "bypass_core0": {
> +            "type": "boolean",
> +            "description": "Indicate that DPDK should omit using the
> first core."
> +          },
> +         "cpus": {
> +            "type": "string",
> +            "description": "Optional comma-separated list of cpus to use,
> e.g.: 1,2,3,4,5,18-22. Defaults to 1. An empty string means use all cpus."
> +          },
> +         "memory_channels": {
> +            "type": "integer",
> +            "description": "How many memory channels to use. Optional,
> defaults to 1."
>            }
>          },
>          "additionalProperties": false,
> @@ -91,6 +112,7 @@
>            "name",
>            "hostname",
>            "user",
> +          "arch",
>            "os"
>          ]
>        },
> diff --git a/dts/framework/remote_session/os/linux_session.py
> b/dts/framework/remote_session/os/linux_session.py
> index 39e80631dd..21f117b714 100644
> --- a/dts/framework/remote_session/os/linux_session.py
> +++ b/dts/framework/remote_session/os/linux_session.py
> @@ -2,6 +2,8 @@
>  # Copyright(c) 2022 PANTHEON.tech s.r.o.
>  # Copyright(c) 2022 University of New Hampshire
>
> +from framework.config import CPU
> +
>  from .posix_session import PosixSession
>
>
> @@ -9,3 +11,16 @@ class LinuxSession(PosixSession):
>      """
>      The implementation of non-Posix compliant parts of Linux remote
> sessions.
>      """
> +
> +    def get_remote_cpus(self, bypass_core0: bool) -> list[CPU]:
> +        cpu_info = self.remote_session.send_command(
> +            "lscpu -p=CPU,CORE,SOCKET,NODE|grep -v \\#"
> +        ).stdout
> +        cpus = []
> +        for cpu_line in cpu_info.splitlines():
> +            cpu, core, socket, node = cpu_line.split(",")
> +            if bypass_core0 and core == 0 and socket == 0:
> +                self.logger.info("Core0 bypassed.")
> +                continue
> +            cpus.append(CPU(int(cpu), int(core), int(socket), int(node)))
> +        return cpus
> diff --git a/dts/framework/remote_session/os/os_session.py
> b/dts/framework/remote_session/os/os_session.py
> index 57e2865282..6f6b6a979e 100644
> --- a/dts/framework/remote_session/os/os_session.py
> +++ b/dts/framework/remote_session/os/os_session.py
> @@ -3,9 +3,10 @@
>  # Copyright(c) 2022 University of New Hampshire
>
>  from abc import ABC, abstractmethod
> +from collections.abc import Iterable
>  from pathlib import PurePath
>
> -from framework.config import Architecture, NodeConfiguration
> +from framework.config import CPU, Architecture, NodeConfiguration
>  from framework.logger import DTSLOG
>  from framework.remote_session.factory import create_remote_session
>  from framework.remote_session.remote_session import RemoteSession
> @@ -130,3 +131,16 @@ def get_dpdk_version(self, version_path: str |
> PurePath) -> str:
>          """
>          Inspect DPDK version on the remote node from version_path.
>          """
> +
> +    @abstractmethod
> +    def get_remote_cpus(self, bypass_core0: bool) -> list[CPU]:
> +        """
> +        Compose a list of CPUs present on the remote node.
> +        """
> +
> +    @abstractmethod
> +    def kill_cleanup_dpdk_apps(self, dpdk_prefix_list: Iterable[str]) ->
> None:
> +        """
> +        Kill and cleanup all DPDK apps identified by dpdk_prefix_list. If
> +        dpdk_prefix_list is empty, attempt to find running DPDK apps to
> kill and clean.
> +        """
> diff --git a/dts/framework/remote_session/os/posix_session.py
> b/dts/framework/remote_session/os/posix_session.py
> index a36b8e8c1a..7151263c7a 100644
> --- a/dts/framework/remote_session/os/posix_session.py
> +++ b/dts/framework/remote_session/os/posix_session.py
> @@ -2,6 +2,8 @@
>  # Copyright(c) 2022 PANTHEON.tech s.r.o.
>  # Copyright(c) 2022 University of New Hampshire
>
> +import re
> +from collections.abc import Iterable
>  from pathlib import PurePath, PurePosixPath
>
>  from framework.config import Architecture
> @@ -138,3 +140,81 @@ def get_dpdk_version(self, build_dir: str | PurePath)
> -> str:
>              f"cat {self.join_remote_path(build_dir, 'VERSION')}",
> verify=True
>          )
>          return out.stdout
> +
> +    def kill_cleanup_dpdk_apps(self, dpdk_prefix_list: Iterable[str]) ->
> None:
> +        self.logger.info("Cleaning up DPDK apps.")
> +        dpdk_runtime_dirs = self._get_dpdk_runtime_dirs(dpdk_prefix_list)
> +        if dpdk_runtime_dirs:
> +            # kill and cleanup only if DPDK is running
> +            dpdk_pids = self._get_dpdk_pids(dpdk_runtime_dirs)
> +            for dpdk_pid in dpdk_pids:
> +                self.remote_session.send_command(f"kill -9 {dpdk_pid}",
> 20)
> +            self._check_dpdk_hugepages(dpdk_runtime_dirs)
> +            self._remove_dpdk_runtime_dirs(dpdk_runtime_dirs)
> +
> +    def _get_dpdk_runtime_dirs(
> +        self, dpdk_prefix_list: Iterable[str]
> +    ) -> list[PurePosixPath]:
> +        prefix = PurePosixPath("/var", "run", "dpdk")
> +        if not dpdk_prefix_list:
> +            remote_prefixes = self._list_remote_dirs(prefix)
> +            if not remote_prefixes:
> +                dpdk_prefix_list = []
> +            else:
> +                dpdk_prefix_list = remote_prefixes
> +
> +        return [PurePosixPath(prefix, dpdk_prefix) for dpdk_prefix in
> dpdk_prefix_list]
> +
> +    def _list_remote_dirs(self, remote_path: str | PurePath) -> list[str]
> | None:
> +        """
> +        Return a list of directories of the remote_dir.
> +        If remote_path doesn't exist, return None.
> +        """
> +        out = self.remote_session.send_command(
> +            f"ls -l {remote_path} | awk '/^d/ {{print $NF}}'"
> +        ).stdout
> +        if "No such file or directory" in out:
> +            return None
> +        else:
> +            return out.splitlines()
> +
> +    def _get_dpdk_pids(self, dpdk_runtime_dirs: Iterable[str | PurePath])
> -> list[int]:
> +        pids = []
> +        pid_regex = r"p(\d+)"
> +        for dpdk_runtime_dir in dpdk_runtime_dirs:
> +            dpdk_config_file = PurePosixPath(dpdk_runtime_dir, "config")
> +            if self._remote_files_exists(dpdk_config_file):
> +                out = self.remote_session.send_command(
> +                    f"lsof -Fp {dpdk_config_file}"
> +                ).stdout
> +                if out and "No such file or directory" not in out:
> +                    for out_line in out.splitlines():
> +                        match = re.match(pid_regex, out_line)
> +                        if match:
> +                            pids.append(int(match.group(1)))
> +        return pids
> +
> +    def _remote_files_exists(self, remote_path: PurePath) -> bool:
> +        result = self.remote_session.send_command(f"test -e
> {remote_path}")
> +        return not result.return_code
> +
> +    def _check_dpdk_hugepages(
> +        self, dpdk_runtime_dirs: Iterable[str | PurePath]
> +    ) -> None:
> +        for dpdk_runtime_dir in dpdk_runtime_dirs:
> +            hugepage_info = PurePosixPath(dpdk_runtime_dir,
> "hugepage_info")
> +            if self._remote_files_exists(hugepage_info):
> +                out = self.remote_session.send_command(
> +                    f"lsof -Fp {hugepage_info}"
> +                ).stdout
> +                if out and "No such file or directory" not in out:
> +                    self.logger.warning("Some DPDK processes did not free
> hugepages.")
> +
> self.logger.warning("*******************************************")
> +                    self.logger.warning(out)
> +
> self.logger.warning("*******************************************")
> +
> +    def _remove_dpdk_runtime_dirs(
> +        self, dpdk_runtime_dirs: Iterable[str | PurePath]
> +    ) -> None:
> +        for dpdk_runtime_dir in dpdk_runtime_dirs:
> +            self.remove_remote_dir(dpdk_runtime_dir)
> diff --git a/dts/framework/testbed_model/hw/__init__.py
> b/dts/framework/testbed_model/hw/__init__.py
> new file mode 100644
> index 0000000000..7d79a7efd0
> --- /dev/null
> +++ b/dts/framework/testbed_model/hw/__init__.py
> @@ -0,0 +1,17 @@
> +# SPDX-License-Identifier: BSD-3-Clause
> +# Copyright(c) 2022 PANTHEON.tech s.r.o.
> +
> +from framework.config import CPU, CPUList
> +
> +from .cpu import CPUAmount, CPUAmountFilter, CPUFilter, CPUListFilter
> +
> +
> +def cpu_filter(
> +    core_list: list[CPU], filter_specifier: CPUAmount | CPUList,
> ascending: bool
> +) -> CPUFilter:
> +    if isinstance(filter_specifier, CPUList):
> +        return CPUListFilter(core_list, filter_specifier, ascending)
> +    elif isinstance(filter_specifier, CPUAmount):
> +        return CPUAmountFilter(core_list, filter_specifier, ascending)
> +    else:
> +        raise ValueError(f"Unsupported filter r{filter_specifier}")
> diff --git a/dts/framework/testbed_model/hw/cpu.py
> b/dts/framework/testbed_model/hw/cpu.py
> new file mode 100644
> index 0000000000..87e87bcb4e
> --- /dev/null
> +++ b/dts/framework/testbed_model/hw/cpu.py
> @@ -0,0 +1,164 @@
> +# SPDX-License-Identifier: BSD-3-Clause
> +# Copyright(c) 2022 PANTHEON.tech s.r.o.
> +
> +import dataclasses
> +from abc import ABC, abstractmethod
> +from collections.abc import Iterable
> +
> +from framework.config import CPU, CPUList
> +
> +
> +@dataclasses.dataclass(slots=True, frozen=True)
> +class CPUAmount:
> +    """
> +    Define the amounts of cpus to use. If sockets is not None,
> socket_amount
> +    is ignored.
> +    """
> +
> +    cpus_per_core: int = 1
> +    cores_per_socket: int = 2
> +    socket_amount: int = 1
> +    sockets: list[int] | None = None
> +
> +
> +class CPUFilter(ABC):
> +    """
> +    Filter according to the input filter specifier. Each filter needs to
> be
> +    implemented in a derived class.
> +    This class only implements operations common to all filters, such as
> sorting
> +    the list to be filtered beforehand.
> +    """
> +
> +    _filter_specifier: CPUAmount | CPUList
> +    _cpus_to_filter: list[CPU]
> +
> +    def __init__(
> +        self,
> +        core_list: list[CPU],
> +        filter_specifier: CPUAmount | CPUList,
> +        ascending: bool = True,
> +    ) -> None:
> +        self._filter_specifier = filter_specifier
> +
> +        # sorting by core is needed in case hyperthreading is enabled
> +        self._cpus_to_filter = sorted(
> +            core_list, key=lambda x: x.core, reverse=not ascending
> +        )
> +        self.filter()
> +
> +    @abstractmethod
> +    def filter(self) -> list[CPU]:
> +        """
> +        Use the input self._filter_specifier to filter
> self._cpus_to_filter
> +        and return the list of filtered CPUs. self._cpus_to_filter is a
> +        sorter copy of the original list, so it may be modified.
> +        """
> +
> +
> +class CPUAmountFilter(CPUFilter):
> +    """
> +    Filter the input list of CPUs according to specified rules:
> +    Use cores from the specified amount of sockets or from the specified
> socket ids.
> +    If sockets is specified, it takes precedence over socket_amount.
> +    From each of those sockets, use only cores_per_socket of cores.
> +    And for each core, use cpus_per_core of cpus. Hypertheading
> +    must be enabled for this to take effect.
> +    If ascending is True, use cores with the lowest numerical id first
> +    and continue in ascending order. If False, start with the highest
> +    id and continue in descending order. This ordering affects which
> +    sockets to consider first as well.
> +    """
> +
> +    _filter_specifier: CPUAmount
> +
> +    def filter(self) -> list[CPU]:
> +        return
> self._filter_cpus(self._filter_sockets(self._cpus_to_filter))
> +
> +    def _filter_sockets(self, cpus_to_filter: Iterable[CPU]) -> list[CPU]:
> +        allowed_sockets: set[int] = set()
> +        socket_amount = self._filter_specifier.socket_amount
> +        if self._filter_specifier.sockets:
> +            socket_amount = len(self._filter_specifier.sockets)
> +            allowed_sockets = set(self._filter_specifier.sockets)
> +
> +        filtered_cpus = []
> +        for cpu in cpus_to_filter:
> +            if not self._filter_specifier.sockets:
> +                if len(allowed_sockets) < socket_amount:
> +                    allowed_sockets.add(cpu.socket)
> +            if cpu.socket in allowed_sockets:
> +                filtered_cpus.append(cpu)
> +
> +        if len(allowed_sockets) < socket_amount:
> +            raise ValueError(
> +                f"The amount of sockets from which to use cores "
> +                f"({socket_amount}) exceeds the actual amount present "
> +                f"on the node ({len(allowed_sockets)})"
> +            )
> +
> +        return filtered_cpus
> +
> +    def _filter_cpus(self, cpus_to_filter: Iterable[CPU]) -> list[CPU]:
> +        # no need to use ordered dict, from Python3.7 the dict
> +        # insertion order is preserved (LIFO).
> +        allowed_cpu_per_core_count_map: dict[int, int] = {}
> +        filtered_cpus = []
> +        for cpu in cpus_to_filter:
> +            if cpu.core in allowed_cpu_per_core_count_map:
> +                cpu_count = allowed_cpu_per_core_count_map[cpu.core]
> +                if self._filter_specifier.cpus_per_core > cpu_count:
> +                    # only add cpus of the given core
> +                    allowed_cpu_per_core_count_map[cpu.core] += 1
> +                    filtered_cpus.append(cpu)
> +                else:
> +                    raise ValueError(
> +                        f"The amount of CPUs per core to use "
> +                        f"({self._filter_specifier.cpus_per_core}) "
> +                        f"exceeds the actual amount present. Is
> hyperthreading enabled?"
> +                    )
> +            elif self._filter_specifier.cores_per_socket > len(
> +                allowed_cpu_per_core_count_map
> +            ):
> +                # only add cpus if we need more
> +                allowed_cpu_per_core_count_map[cpu.core] = 1
> +                filtered_cpus.append(cpu)
> +            else:
> +                # cpus are sorted by core, at this point we won't
> encounter new cores
> +                break
> +
> +        cores_per_socket = len(allowed_cpu_per_core_count_map)
> +        if cores_per_socket < self._filter_specifier.cores_per_socket:
> +            raise ValueError(
> +                f"The amount of cores per socket to use "
> +                f"({self._filter_specifier.cores_per_socket}) "
> +                f"exceeds the actual amount present ({cores_per_socket})"
> +            )
> +
> +        return filtered_cpus
> +
> +
> +class CPUListFilter(CPUFilter):
> +    """
> +    Filter the input list of CPUs according to the input list of
> +    core indices.
> +    An empty CPUList won't filter anything.
> +    """
> +
> +    _filter_specifier: CPUList
> +
> +    def filter(self) -> list[CPU]:
> +        if not len(self._filter_specifier.cpu_list):
> +            return self._cpus_to_filter
> +
> +        filtered_cpus = []
> +        for core in self._cpus_to_filter:
> +            if core.cpu in self._filter_specifier.cpu_list:
> +                filtered_cpus.append(core)
> +
> +        if len(filtered_cpus) != len(self._filter_specifier.cpu_list):
> +            raise ValueError(
> +                f"Not all cpus from {self._filter_specifier.cpu_list}
> were found"
> +                f"among {self._cpus_to_filter}"
> +            )
> +
> +        return filtered_cpus
> diff --git a/dts/framework/testbed_model/node/node.py
> b/dts/framework/testbed_model/node/node.py
> index 86654e55ae..5ee7023335 100644
> --- a/dts/framework/testbed_model/node/node.py
> +++ b/dts/framework/testbed_model/node/node.py
> @@ -8,13 +8,16 @@
>  """
>
>  from framework.config import (
> +    CPU,
>      BuildTargetConfiguration,
> +    CPUList,
>      ExecutionConfiguration,
>      NodeConfiguration,
>  )
>  from framework.exception import NodeCleanupError, NodeSetupError,
> convert_exception
>  from framework.logger import DTSLOG, getLogger
>  from framework.remote_session import OSSession, create_session
> +from framework.testbed_model.hw import CPUAmount, cpu_filter
>
>
>  class Node(object):
> @@ -28,6 +31,7 @@ class Node(object):
>      main_session: OSSession
>      logger: DTSLOG
>      config: NodeConfiguration
> +    cpus: list[CPU]
>      _other_sessions: list[OSSession]
>
>      def __init__(self, node_config: NodeConfiguration):
> @@ -38,6 +42,7 @@ def __init__(self, node_config: NodeConfiguration):
>          self.logger = getLogger(self.name)
>          self.logger.info(f"Created node: {self.name}")
>          self.main_session = create_session(self.config, self.name,
> self.logger)
> +        self._get_remote_cpus()
>
>      @convert_exception(NodeSetupError)
>      def setup_execution(self, execution_config: ExecutionConfiguration)
> -> None:
> @@ -109,6 +114,37 @@ def create_session(self, name: str) -> OSSession:
>          self._other_sessions.append(connection)
>          return connection
>
> +    def filter_cpus(
> +        self,
> +        filter_specifier: CPUAmount | CPUList,
> +        ascending: bool = True,
> +    ) -> list[CPU]:
> +        """
> +        Filter the logical cpus found on the Node according to specified
> rules:
> +        Use cores from the specified amount of sockets or from the
> specified
> +        socket ids. If sockets is specified, it takes precedence over
> socket_amount.
> +        From each of those sockets, use only cpus_per_socket of cores.
> +        And for each core, use cpus_per_core of cpus. Hypertheading
> +        must be enabled for this to take effect.
> +        If ascending is True, use cores with the lowest numerical id first
> +        and continue in ascending order. If False, start with the highest
> +        id and continue in descending order. This ordering affects which
> +        sockets to consider first as well.
> +        """
> +        self.logger.info("Filtering ")
> +        return cpu_filter(
> +            self.cpus,
> +            filter_specifier,
> +            ascending,
> +        ).filter()
> +
> +    def _get_remote_cpus(self) -> None:
> +        """
> +        Scan cpus in the remote OS and store a list of CPUs.
> +        """
> +        self.logger.info("Getting CPU information.")
> +        self.cpus =
> self.main_session.get_remote_cpus(self.config.bypass_core0)
> +
>      def close(self) -> None:
>          """
>          Close all connections and free other resources.
> diff --git a/dts/framework/testbed_model/node/sut_node.py
> b/dts/framework/testbed_model/node/sut_node.py
> index 53268a7565..ff3be845b4 100644
> --- a/dts/framework/testbed_model/node/sut_node.py
> +++ b/dts/framework/testbed_model/node/sut_node.py
> @@ -4,10 +4,13 @@
>
>  import os
>  import tarfile
> +import time
>  from pathlib import PurePath
>
> -from framework.config import BuildTargetConfiguration, NodeConfiguration
> +from framework.config import CPU, BuildTargetConfiguration, CPUList,
> NodeConfiguration
> +from framework.remote_session import OSSession
>  from framework.settings import SETTINGS
> +from framework.testbed_model.hw import CPUAmount, CPUListFilter
>  from framework.utils import EnvVarsDict, skip_setup
>
>  from .node import Node
> @@ -21,19 +24,31 @@ class SutNode(Node):
>      Another key capability is building DPDK according to given build
> target.
>      """
>
> +    cpus: list[CPU]
> +    dpdk_prefix_list: list[str]
> +    dpdk_prefix_subfix: str
>      _build_target_config: BuildTargetConfiguration | None
>      _env_vars: EnvVarsDict
>      _remote_tmp_dir: PurePath
>      __remote_dpdk_dir: PurePath | None
>      _app_compile_timeout: float
> +    _dpdk_kill_session: OSSession | None
>
>      def __init__(self, node_config: NodeConfiguration):
>          super(SutNode, self).__init__(node_config)
> +        self.dpdk_prefix_list = []
>          self._build_target_config = None
>          self._env_vars = EnvVarsDict()
>          self._remote_tmp_dir = self.main_session.get_remote_tmp_dir()
>          self.__remote_dpdk_dir = None
>          self._app_compile_timeout = 90
> +        self._dpdk_kill_session = None
> +
> +        # filter the node cpus according to user config
> +        self.cpus = CPUListFilter(self.cpus, self.config.cpus).filter()
> +        self.dpdk_prefix_subfix = (
> +            f"{str(os.getpid())}_{time.strftime('%Y%m%d%H%M%S',
> time.localtime())}"
> +        )
>
>      @property
>      def _remote_dpdk_dir(self) -> PurePath:
> @@ -142,3 +157,164 @@ def build_dpdk_app(self, app_name: str) -> PurePath:
>          return self.main_session.join_remote_path(
>              build_dir, "examples", f"dpdk-{app_name}"
>          )
> +
> +    def kill_cleanup_dpdk_apps(self) -> None:
> +        """
> +        Kill all dpdk applications on the SUT. Cleanup hugepages.
> +        """
> +        if self._dpdk_kill_session and self._dpdk_kill_session.is_alive():
> +            # we can use the session if it exists and responds
> +
> self._dpdk_kill_session.kill_cleanup_dpdk_apps(self.dpdk_prefix_list)
> +        else:
> +            # otherwise, we need to (re)create it
> +            self._dpdk_kill_session = self.create_session("dpdk_kill")
> +        self.dpdk_prefix_list = []
> +
> +    def create_eal_parameters(
> +        self,
> +        fixed_prefix: bool = False,
> +        core_filter_specifier: CPUAmount | CPUList = CPUAmount(),
> +        ascending_cores: bool = True,
> +        prefix: str = "",
> +        no_pci: bool = False,
> +        vdevs: list[str] = None,
>

I would prefer to have vdevs be a list of objects, even if for now that
class just takes a string in its constructor. Later on we can add
subclasses for specific vdevs that might see heavy use, such
as librte_net_pcap and crypto_openssl.


> +        other_eal_param: str = "",
> +    ) -> str:
> +        """
> +        Generate eal parameters character string;
> +        :param fixed_prefix: use fixed file-prefix or not, when it is
> true,
> +                        the file-prefix will not be added a timestamp
> +        :param core_filter_specifier: an amount of cpus/cores/sockets to
> use
> +                        or a list of cpu ids to use.
> +                        The default will select one cpu for each of two
> cores
> +                        on one socket, in ascending order of core ids.
> +        :param ascending_cores: True, use cores with the lowest numerical
> id first
> +                        and continue in ascending order. If False, start
> with the
> +                        highest id and continue in descending order. This
> ordering
> +                        affects which sockets to consider first as well.
> +        :param prefix: set file prefix string, eg:
> +                        prefix='vf';
> +        :param no_pci: switch of disable PCI bus eg:
> +                        no_pci=True;
> +        :param vdevs: virtual device list, eg:
> +                        vdevs=['net_ring0', 'net_ring1'];
> +        :param other_eal_param: user defined DPDK eal parameters, eg:
> +                        other_eal_param='--single-file-segments';
> +        :return: eal param string, eg:
> +                '-c 0xf -a 0000:88:00.0
> --file-prefix=dpdk_1112_20190809143420';
> +        if DPDK version < 20.11-rc4, eal_str eg:
> +                '-c 0xf -w 0000:88:00.0
> --file-prefix=dpdk_1112_20190809143420';
> +        """
> +        if vdevs is None:
> +            vdevs = []
> +
> +        config = {
> +            "core_filter_specifier": core_filter_specifier,
> +            "ascending_cores": ascending_cores,
> +            "prefix": prefix,
> +            "no_pci": no_pci,
> +            "vdevs": vdevs,
> +            "other_eal_param": other_eal_param,
> +        }
> +
> +        eal_parameter_creator = _EalParameter(
> +            sut_node=self, fixed_prefix=fixed_prefix, **config
> +        )
> +        eal_str = eal_parameter_creator.make_eal_param()
> +
> +        return eal_str
> +
> +
> +class _EalParameter(object):
> +    def __init__(
> +        self,
> +        sut_node: SutNode,
> +        fixed_prefix: bool,
> +        core_filter_specifier: CPUAmount | CPUList,
> +        ascending_cores: bool,
> +        prefix: str,
> +        no_pci: bool,
> +        vdevs: list[str],
> +        other_eal_param: str,
> +    ):
> +        """
> +        Generate eal parameters character string;
> +        :param sut_node: SUT Node;
> +        :param fixed_prefix: use fixed file-prefix or not, when it is
> true,
> +                        he file-prefix will not be added a timestamp
> +        :param core_filter_specifier: an amount of cpus/cores/sockets to
> use
> +                        or a list of cpu ids to use.
> +        :param ascending_cores: True, use cores with the lowest numerical
> id first
> +                        and continue in ascending order. If False, start
> with the
> +                        highest id and continue in descending order. This
> ordering
> +                        affects which sockets to consider first as well.
> +        :param prefix: set file prefix string, eg:
> +                        prefix='vf';
> +        :param no_pci: switch of disable PCI bus eg:
> +                        no_pci=True;
> +        :param vdevs: virtual device list, eg:
> +                        vdevs=['net_ring0', 'net_ring1'];
> +        :param other_eal_param: user defined DPDK eal parameters, eg:
> +                        other_eal_param='--single-file-segments';
> +        """
> +        self.os = sut_node.config.os
> +        self.fixed_prefix = fixed_prefix
> +        self.sut_node = sut_node
> +        self.core_filter_specifier = core_filter_specifier
> +        self.ascending_cores = ascending_cores
> +        self.prefix = prefix
> +        self.no_pci = no_pci
> +        self.vdevs = vdevs
> +        self.other_eal_param = other_eal_param
> +
> +    def _make_lcores_param(self) -> str:
> +        filtered_cpus = self.sut_node.filter_cpus(
> +            self.core_filter_specifier, self.ascending_cores
> +        )
> +        return f"-l {CPUList(filtered_cpus)}"
> +
> +    def _make_memory_channels(self) -> str:
> +        param_template = "-n {}"
> +        return param_template.format(self.sut_node.config.memory_channels)
> +
> +    def _make_no_pci_param(self) -> str:
> +        if self.no_pci is True:
> +            return "--no-pci"
> +        else:
> +            return ""
> +
> +    def _make_prefix_param(self) -> str:
> +        if self.prefix == "":
> +            fixed_file_prefix = f"dpdk_{self.sut_node.dpdk_prefix_subfix}"
> +        else:
> +            fixed_file_prefix = self.prefix
> +            if not self.fixed_prefix:
> +                fixed_file_prefix = (
> +
> f"{fixed_file_prefix}_{self.sut_node.dpdk_prefix_subfix}"
> +                )
> +        fixed_file_prefix =
> self._do_os_handle_with_prefix_param(fixed_file_prefix)
> +        return fixed_file_prefix
> +
> +    def _make_vdevs_param(self) -> str:
> +        if len(self.vdevs) == 0:
> +            return ""
> +        else:
> +            return " ".join(f"--vdev {vdev}" for vdev in self.vdevs)
> +
> +    def _do_os_handle_with_prefix_param(self, file_prefix: str) -> str:
> +        self.sut_node.dpdk_prefix_list.append(file_prefix)
> +        return f"--file-prefix={file_prefix}"
> +
> +    def make_eal_param(self) -> str:
> +        _eal_str = " ".join(
> +            [
> +                self._make_lcores_param(),
> +                self._make_memory_channels(),
> +                self._make_prefix_param(),
> +                self._make_no_pci_param(),
> +                self._make_vdevs_param(),
> +                # append user defined eal parameters
> +                self.other_eal_param,
> +            ]
> +        )
> +        return _eal_str
> diff --git a/dts/framework/utils.py b/dts/framework/utils.py
> index 91e58f3218..3c2f0adff9 100644
> --- a/dts/framework/utils.py
> +++ b/dts/framework/utils.py
> @@ -32,6 +32,26 @@ def skip_setup(func) -> Callable[..., None]:
>          return func
>
>
> +def expand_range(range_str: str) -> list[int]:
> +    """
> +    Process range string into a list of integers. There are two possible
> formats:
> +    n - a single integer
> +    n-m - a range of integers
> +
> +    The returned range includes both n and m. Empty string returns an
> empty list.
> +    """
> +    expanded_range: list[int] = []
> +    if range_str:
> +        range_boundaries = range_str.split("-")
> +        # will throw an exception when items in range_boundaries can't be
> converted,
> +        # serving as type check
> +        expanded_range.extend(
> +            range(int(range_boundaries[0]), int(range_boundaries[-1]) + 1)
> +        )
> +
> +    return expanded_range
> +
> +
>  def GREEN(text: str) -> str:
>      return f"\u001B[32;1m{str(text)}\u001B[0m"
>
> --
> 2.30.2
>
>
  

Again, apologies for removing recipients in my earlier reply.

From: Owen Hilyard <ohilyard@iol.unh.edu>
Sent: Monday, November 21, 2022 1:40 PM
To: Juraj Linkeš <juraj.linkes@pantheon.tech>
Subject: Re: [RFC PATCH v2 04/10] dts: add dpdk execution handling

On Fri, Nov 18, 2022 at 8:00 AM Juraj Linkeš <juraj.linkes@pantheon.tech<mailto:juraj.linkes@pantheon.tech>> wrote:
diff --git a/dts/framework/config/conf_yaml_schema.json b/dts/framework/config/conf_yaml_schema.json
index 409ce7ac74..c59d3e30e6 100644
--- a/dts/framework/config/conf_yaml_schema.json
+++ b/dts/framework/config/conf_yaml_schema.json
@@ -6,6 +6,12 @@
       "type": "string",
       "description": "A unique identifier for a node"
     },
+    "ARCH": {
+      "type": "string",
+      "enum": [
+        "x86_64"

arm64 and ppc64le should probably be included here. I think that we can focus on 64 bit arches for now.

[Juraj] Seems safe enough. At this point it doesn't matter, but when we have a number of testcases, we may need to revisit this (if we can't verify an architecture for example).

[Owen] The reason I want this is because I want there to always be an architecture that is not the one being developed on that developers need to handle properly. LoongArch might actually be a good candidate for this if support gets merged, since to my knowledge almost no one has access to their server-class CPUs yet. Essentially, I want to force anyone who does something that is architecture dependent to consider other architectures, not just have the "the entire world is x86" mentality.

Alright, good to know.
I have a semi-related point, we specify arch (and os as well) in both build target and SUT config. Are these even going to be different? I see cpu (or platform in meson config) being different, but not the other two and that could simplify the config a bit.

<snip>

+    def kill_cleanup_dpdk_apps(self) -> None:
+        """
+        Kill all dpdk applications on the SUT. Cleanup hugepages.
+        """
+        if self._dpdk_kill_session and self._dpdk_kill_session.is_alive():
+            # we can use the session if it exists and responds
+            self._dpdk_kill_session.kill_cleanup_dpdk_apps(self.dpdk_prefix_list)
+        else:
+            # otherwise, we need to (re)create it
+            self._dpdk_kill_session = self.create_session("dpdk_kill")
+        self.dpdk_prefix_list = []
+
+    def create_eal_parameters(
+        self,
+        fixed_prefix: bool = False,
+        core_filter_specifier: CPUAmount | CPUList = CPUAmount(),
+        ascending_cores: bool = True,
+        prefix: str = "",
+        no_pci: bool = False,
+        vdevs: list[str] = None,

I would prefer to have vdevs be a list of objects, even if for now that class just takes a string in its constructor. Later on we can add subclasses for specific vdevs that might see heavy use, such as librte_net_pcap and crypto_openssl.

[Juraj] Ok, this is simple enough, I'll add it.
  

On Wed, Nov 23, 2022 at 8:03 AM Juraj Linkeš <juraj.linkes@pantheon.tech>
wrote:

> Again, apologies for removing recipients in my earlier reply.
>
>
>
> *From:* Owen Hilyard <ohilyard@iol.unh.edu>
> *Sent:* Monday, November 21, 2022 1:40 PM
> *To:* Juraj Linkeš <juraj.linkes@pantheon.tech>
> *Subject:* Re: [RFC PATCH v2 04/10] dts: add dpdk execution handling
>
>
>
> On Fri, Nov 18, 2022 at 8:00 AM Juraj Linkeš <juraj.linkes@pantheon.tech>
> wrote:
>
> diff --git a/dts/framework/config/conf_yaml_schema.json
> b/dts/framework/config/conf_yaml_schema.json
> index 409ce7ac74..c59d3e30e6 100644
> --- a/dts/framework/config/conf_yaml_schema.json
> +++ b/dts/framework/config/conf_yaml_schema.json
> @@ -6,6 +6,12 @@
>        "type": "string",
>        "description": "A unique identifier for a node"
>      },
> +    "ARCH": {
> +      "type": "string",
> +      "enum": [
> +        "x86_64"
>
> arm64 and ppc64le should probably be included here. I think that we can
> focus on 64 bit arches for now.
>
> [Juraj] Seems safe enough. At this point it doesn't matter, but when we
> have a number of testcases, we may need to revisit this (if we can't verify
> an architecture for example).
>
>
>
> [Owen] The reason I want this is because I want there to always be an
> architecture that is not the one being developed on that developers need to
> handle properly. LoongArch might actually be a good candidate for this if
> support gets merged, since to my knowledge almost no one has access to
> their server-class CPUs yet. Essentially, I want to force anyone who does
> something that is architecture dependent to consider other architectures,
> not just have the "the entire world is x86" mentality.
>
>
>
> Alright, good to know.
>
> I have a semi-related point, we specify arch (and os as well) in both
> build target and SUT config. Are these even going to be different? I see
> cpu (or platform in meson config) being different, but not the other two
> and that could simplify the config a bit.
>

[Owen] If I remember correctly, the older DTS has i686 (32 bit x86)
support, and you might want to run i686 on an x86_64 cpu. That is the only
use case I can see for differing build arch and SUT arch. The community lab
doesn't have any 32 bit hardware, so any future 32 bit testing would need
to happen on a 64 bit system running in a compatibility mode.


> <snip>
>
> +    def kill_cleanup_dpdk_apps(self) -> None:
> +        """
> +        Kill all dpdk applications on the SUT. Cleanup hugepages.
> +        """
> +        if self._dpdk_kill_session and self._dpdk_kill_session.is_alive():
> +            # we can use the session if it exists and responds
> +
> self._dpdk_kill_session.kill_cleanup_dpdk_apps(self.dpdk_prefix_list)
> +        else:
> +            # otherwise, we need to (re)create it
> +            self._dpdk_kill_session = self.create_session("dpdk_kill")
> +        self.dpdk_prefix_list = []
> +
> +    def create_eal_parameters(
> +        self,
> +        fixed_prefix: bool = False,
> +        core_filter_specifier: CPUAmount | CPUList = CPUAmount(),
> +        ascending_cores: bool = True,
> +        prefix: str = "",
> +        no_pci: bool = False,
> +        vdevs: list[str] = None,
>
> I would prefer to have vdevs be a list of objects, even if for now that
> class just takes a string in its constructor. Later on we can add
> subclasses for specific vdevs that might see heavy use, such
> as librte_net_pcap and crypto_openssl.
>
> [Juraj] Ok, this is simple enough, I'll add it.
>
>