[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <91cba75f-0997-43e8-93d0-b795b3783eff@proton.me>
Date: Sat, 28 Oct 2023 16:39:08 +0000
From: Benno Lossin <benno.lossin@...ton.me>
To: FUJITA Tomonori <fujita.tomonori@...il.com>, andrew@...n.ch
Cc: boqun.feng@...il.com, netdev@...r.kernel.org, rust-for-linux@...r.kernel.org, tmgross@...ch.edu, miguel.ojeda.sandonis@...il.com, wedsonaf@...il.com
Subject: Re: [PATCH net-next v7 1/5] rust: core abstractions for network PHY drivers
On 28.10.23 18:09, FUJITA Tomonori wrote:
> On Sat, 28 Oct 2023 16:53:30 +0200
> Andrew Lunn <andrew@...n.ch> wrote:
>
>>>> We need to be careful here, since doing this creates a reference
>>>> `&bindings::phy_device` which asserts that it is immutable. That is not
>>>> the case, since the C side might change it at any point (this is the
>>>> reason we wrap things in `Opaque`, since that allows mutatation even
>>>> through sharde references).
>>>
>>> You meant that the C code might modify it independently anytime, not
>>> the C code called the Rust abstractions might modify it, right?
>>
>> The whole locking model is base around that not happening. Things
>> should only change with the lock held. I you make a call into the C
>> side, then yes, it can and will change it. So you should not cache a
>> value over a C call.
>
> Yeah, I understand that. But if I understand Benno correctly, from
> Rust perspective, such might happen.
Yes, that is what I meant. Sure the C side might never modify the
value, but this is not good enough for Rust. It must somehow be ensured
that it never is modified, in order for us to rely on it.
--
Cheers,
Benno
Powered by blists - more mailing lists