From: "Juraj Linkeš" <juraj.linkes@pantheon.tech>
To: Luca Vizzarro <Luca.Vizzarro@arm.com>
Cc: Jeremy Spewock <jspewock@iol.unh.edu>,
dev@dpdk.org, Jack Bond-Preston <jack.bond-preston@arm.com>,
Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Subject: Re: [PATCH 6/6] dts: add statefulness to TestPmdShell
Date: Thu, 11 Apr 2024 14:13:34 +0200 [thread overview]
Message-ID: <CAOb5WZbjdDBg5rL-_JQKxt-J0BfUjEEhp9LtZTtEq3M+mzEhFw@mail.gmail.com> (raw)
In-Reply-To: <5d35012d-0ffd-4c23-ad0c-8315453b8c9e@arm.com>
On Thu, Apr 11, 2024 at 1:47 PM Luca Vizzarro <Luca.Vizzarro@arm.com> wrote:
>
> On 11/04/2024 11:30, Juraj Linkeš wrote:
> > I've been thinking about these interactive shell constructors for some
> > time and I think the factory pattern is not well suitable for this.
> > Factories work well with classes with the same API (i.e.
> > implementations of abstract classes that don't add anything extra),
> > but are much less useful when dealing with classes with different
> > behaviors, such as the interactive shells. We see this here, different
> > apps are going to require different args and that alone kinda breaks
> > the factory pattern. I think we'll need to either ditch these
> > factories and instead just have methods that return the proper shell
> > (and the methods would only exist in classes where they belong, e.g.
> > testpmd only makes sense on an SUT). Or we could overload each factory
> > (the support has only been added in 3.11 with @typing.overload, but is
> > also available in typing_extensions, so we would be able to use it
> > with the extra dependency) where different signatures would return
> > different objects. In both cases the caller won't have to import the
> > class and the method signature is going to be clearer.
> >
> > We have this pattern with sut/tg nodes. I decided to move away from
> > the node factory because it didn't add much and in fact the code was
> > only clunkier. The interactive shell is not quite the same, as the
> > shells are not standalone in the same way the nodes are (the shells
> > are tied to nodes). Let me know what you think about all this - both
> > Luca and Jeremy.
>
> When writing this series, I went down the path of creating a
> `create_testpmd_shell` method at some point as a solution to these
> problems. Realising after that it may be too big of a change, and
> possibly best left to a discussion exactly like this one.
>
The changes we discuss below don't seem that big. What do you think,
do we just add another patch to the series?
> Generics used at this level may be a bit too much, especially for
> Python, as support is not *that* great. I am of the opinion that having
> a dedicated wrapper is easier for the developer and the user. Generics
> are not needed to this level anyways, as we have a limited selection of
> shells that are actually going to be used.
>
> We can also swap the wrapping process to simplify things, instead of:
>
> shell = self.sut_node.create_interactive_shell(TestPmdShell, ..)
>
> do:
>
> shell = TestPmdShell(self.sut_node, ..)
>
> Let the Shell class ingest the node, and not the other way round.
>
I thought about this a bit as well, it's a good approach. The current
design is top-down, as you say, in that "I have a node and I do things
with the node, including starting testpmd on the node". But it could
also be "I have a node, but I also have other non-node resources at my
disposal and it's up to me how I utilize those". If we can make the
imports work then this is likely the best option.
> The current approach appears to me to be top-down instead of bottom-up.
> We take the most abstracted part and we work our way down. But all we
> want is concreteness to the end user (developer).
>
> > Let me illustrate this on the TestPmdShell __init__() method I had in mind:
> >
> > def __init__(self, interactive_session: SSHClient,
> > logger: DTSLogger,
> > get_privileged_command: Callable[[str], str] | None,
> > app_args: EalTestPmdParams | None = None,
> > timeout: float = SETTINGS.timeout,
> > ) -> None:
> > super().__init__(interactive_session, logger, get_privileged_command)
> > self.state = TestPmdState()
> >
> > Where EalTestPmdParams would be something that enforces that
> > app_args.app_params is of the TestPmdParameters type.
> >
> > But thinking more about this, we're probably better off switching the
> > params composition. Instead of TestPmdParameters being part of
> > EalParameters, we do it the other way around. This way the type of
> > app_args could just be TestPmdParameters and the types should work.
> > Or we pass the args separately, but that would likely require ditching
> > the factories and replacing them with methods (or overloading them).
> >
> > And hopefully the imports won't be impossible to solve. :-)
>
> It is what I feared, and I think it may become even more convoluted. As
> you said, ditching the factories will simplify things and make it more
> straightforward. So, we wouldn't find ourselves in problems like these.
>
> I don't have a strong preference in approach between:
> * overloading node methods
> * dedicated node methods
> * let the shells ingest nodes instead
>
> But if I were to give priority, I'd take it from last to first. Letting
> shells ingest nodes will decouple the situation adding an extra step of
> simplification.
+1 for simplification.
> I may not see the full picture though. The two are
> reasonable but, having a dedicated node method will stop the requirement
> to import the shell we need, and it's pretty much equivalent... but
> overloading also is very new to Python, so I may prefer to stick to more
> established.
>
Let's try shells ingesting nodes if the imports work out then. If not,
we can fall back to dedicated node methods.
> Letting TestPmdParams take EalParams, instead of the other way around,
> would naturally follow the bottom-up approach too. Allowing Params to
> arbitrarily append string arguments – as proposed, would also allow
> users to use a plain (EalParams + string). So sounds like a good
> approach overall.
next prev parent reply other threads:[~2024-04-11 12:13 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-26 19:04 [PATCH 0/6] dts: add testpmd params and statefulness Luca Vizzarro
2024-03-26 19:04 ` [PATCH 1/6] dts: add parameters data structure Luca Vizzarro
2024-03-28 16:48 ` Jeremy Spewock
2024-04-09 15:52 ` Luca Vizzarro
2024-04-09 12:10 ` Juraj Linkeš
2024-04-09 16:28 ` Luca Vizzarro
2024-04-10 9:15 ` Juraj Linkeš
2024-04-10 9:51 ` Luca Vizzarro
2024-04-10 10:04 ` Juraj Linkeš
2024-03-26 19:04 ` [PATCH 2/6] dts: use Params for interactive shells Luca Vizzarro
2024-03-28 16:48 ` Jeremy Spewock
2024-04-09 14:56 ` Juraj Linkeš
2024-04-10 9:34 ` Luca Vizzarro
2024-04-10 9:58 ` Juraj Linkeš
2024-03-26 19:04 ` [PATCH 3/6] dts: add testpmd shell params Luca Vizzarro
2024-03-28 16:48 ` Jeremy Spewock
2024-04-09 16:37 ` Juraj Linkeš
2024-04-10 10:49 ` Luca Vizzarro
2024-04-10 13:17 ` Juraj Linkeš
2024-03-26 19:04 ` [PATCH 4/6] dts: use testpmd params for scatter test suite Luca Vizzarro
2024-04-09 19:12 ` Juraj Linkeš
2024-04-10 10:53 ` Luca Vizzarro
2024-04-10 13:18 ` Juraj Linkeš
2024-04-26 18:06 ` Jeremy Spewock
2024-04-29 7:45 ` Juraj Linkeš
2024-03-26 19:04 ` [PATCH 5/6] dts: add statefulness to InteractiveShell Luca Vizzarro
2024-03-28 16:48 ` Jeremy Spewock
2024-04-10 6:53 ` Juraj Linkeš
2024-04-10 11:27 ` Luca Vizzarro
2024-04-10 13:35 ` Juraj Linkeš
2024-04-10 14:07 ` Luca Vizzarro
2024-04-12 12:33 ` Juraj Linkeš
2024-04-29 14:48 ` Jeremy Spewock
2024-03-26 19:04 ` [PATCH 6/6] dts: add statefulness to TestPmdShell Luca Vizzarro
2024-03-28 16:48 ` Jeremy Spewock
2024-04-10 7:41 ` Juraj Linkeš
2024-04-10 11:35 ` Luca Vizzarro
2024-04-11 10:30 ` Juraj Linkeš
2024-04-11 11:47 ` Luca Vizzarro
2024-04-11 12:13 ` Juraj Linkeš [this message]
2024-04-11 13:59 ` Luca Vizzarro
2024-04-26 18:06 ` Jeremy Spewock
2024-04-29 12:06 ` Juraj Linkeš
2024-04-10 7:50 ` Juraj Linkeš
2024-04-10 11:37 ` Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 0/8] dts: add testpmd params Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 1/8] dts: add params manipulation module Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 2/8] dts: use Params for interactive shells Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 3/8] dts: refactor EalParams Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 4/8] dts: remove module-wide imports Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 5/8] dts: add testpmd shell params Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 6/8] dts: use testpmd params for scatter test suite Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 7/8] dts: rework interactive shells Luca Vizzarro
2024-05-09 11:20 ` [PATCH v2 8/8] dts: use Unpack for type checking and hinting Luca Vizzarro
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAOb5WZbjdDBg5rL-_JQKxt-J0BfUjEEhp9LtZTtEq3M+mzEhFw@mail.gmail.com \
--to=juraj.linkes@pantheon.tech \
--cc=Luca.Vizzarro@arm.com \
--cc=dev@dpdk.org \
--cc=honnappa.nagarahalli@arm.com \
--cc=jack.bond-preston@arm.com \
--cc=jspewock@iol.unh.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).