Disable cfg check for the recently-merged target features to allow stdarch update

[sayantn]

The stdarch update is currently giving doc errors, which #1767 fixes. But then it will give unknown cfg errors due to the recently added amx and riscv target features, so this pr temporarily disables those checks. Revert after the bump is successful

[rustbot]

r? @Amanieu

rustbot has assigned @Amanieu.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[a4lg]

@sayantn
I'd like to ask how do we know that unknown cfg errors are fixed (and/or which CI failures to track) to know the right timing to revert this PR.
Just making documentation seems working on my local environment so I think there is a failing configuration.

[sayantn]

These unknown cfg errors are obtained when stdarch is built as a part of libstd, by itself stdarch should compile fine - the bootstrapping architecture of rustc creates the problem because the stage0 compiler doesn't yet know about those new target features.

It should be good to revert once a stdarch update gets merged, or the bootstrap version gets bumped

[Urgau]

It should be good to revert once a stdarch update gets merged, or the bootstrap version gets bumped

I think this could be handled by the special bootstrap cfg set by the rust-lang/rust build-system.

[sayantn]

afaik this repo doesn't use cfg(bootstrap) because the update cfgs commit (which removes all cfg(bootstrap) items after the bootstrap bump) wouldn't work here because this is still a submodule, not a subtree. Using cfg(bootstrap) here will require a PR here after bootstrap bump, and a stdarch submodule update in rustc, which might create a lot of problems, because stdarch updates are typically problematic as stdarch CI and rustc CI don't match perfectly