Remove lifetime from BufferSlice, BufferView, BufferViewMut

[sagudev]

Connections
Continuation of #7764
Related issue: #6974

Description
Currently, BufferView* is generic over a lifetime, but the lifetime is only used for a reference to a buffer. Since buffers are now Cloneable, it make sense to abandon this and instead have BufferView* "own" a Buffer.

This makes it easier to pass around BufferView*.

Testing
Tested via existing tests.

Squash or Rebase?

Squash

Checklist

• Run cargo fmt.
• Run taplo format.
• Run cargo clippy --tests. If applicable, add:
  • --target wasm32-unknown-unknown
• Run cargo xtask test to run tests.
• If this contains user-facing changes, add a CHANGELOG.md entry.

[sagudev]

Using BufferSlice instead of &BufferSlice would be less diff, but less flexible (user would need to clone BufferSlice in some cases).

[kpreid]

The thing I really like about this change is that BufferView[Mut] (the one that gives access to mapped buffers, not just designating a part of a buffer) now has no lifetime, so it can be used in more ways. (But that could also be done on its own without modifying BufferSlice.)

The thing I don't like is that every render pass user has to add more &s, and their .slice() calls now perform a completely unnecessary reference count increment and decrement. Is there any chance it would be possible to arrange so that the set_vertex_buffer() etc. will actually take ownership of the slice's reference count and take advantage of that internally for their own resource tracking purposes? Or are there too many layers in the way? If this were possible, it would keep the syntax for users simpler and be a net performance improvement.

[sagudev]

The thing I don't like is that every render pass user has to add more &s, and their .slice() calls now perform a completely unnecessary reference count increment and decrement. Is there any chance it would be possible to arrange so that the set_vertex_buffer() etc. will actually take ownership of the slice's reference count and take advantage of that internally for their own resource tracking purposes? Or are there too many layers in the way? If this were possible, it would keep the syntax for users simpler and be a net performance improvement.

It might be possible, but that would be more involved as inner method takes &DispathcBuffer anyway.

The thing I really like about this change is that BufferView[Mut] (the one that gives access to mapped buffers, not just designating a part of a buffer) now has no lifetime, so it can be used in more ways. (But that could also be done on its own without modifying BufferSlice.)

I just check and I only need BufferView without lifetimes, so I went with this approach.

[Wumpf]

yay, finally!
This will fix some awful unsafe code I maintain in a different project! :)