Why would he do that? He has no control over an independent library outside of C DMA? I thought he said he's fine with any drivers written in Rust. Just that the maintenance of such including any wrapper cannot fall into C DMA's lap.

> Why would he do that?

It's a great question. I mean, my read of it is he hates the idea of Rust4Linux and is using his position to obstruct.

> He has no control over an independent library outside of C DMA?

Apparently not.

> Just that the maintenance of such including any wrapper cannot fall into C DMA's lap.

The patch he rejected did not add any code to C DMA, nor C DMA's directory (kernel/dma). Just:

  rust/bindings/bindings_helper.h |   1 +
  rust/kernel/dma.rs              | 271 ++++++++++++++++++++++++++++++++
  rust/kernel/lib.rs              |   1 +

(Nor does any Rust4Linux code add any maintenance burden to C -- C maintainers are allowed to break Rust code at will.)

This isn't even just your read of it; as other people in this thread have pointed out, he's pretty explicitly stated it at this point.

No, he said very explicitly that he is against any Rust code in Linux whatsoever. He doesn't have any beef with Rust, the same would be true for him if people tried to add C++ or Pascal or Ada or any other kind of drivers or infra in Linux: he believes that the project will be much harder to maintain if it's not 100% pure C (and assembly).

Here is a link to this explicit claim: https://lore.kernel.org/rust-for-linux/20250131075751.GA1672...

> Why would he do that? He has no control over an independent library outside of C DMA?

That is the crux of the entire drama and why R4L developers got upset at Christoph in the first place, and asked Linus to intervene.