fix(docs): tracking issue for gpu_launch_sized_workgroup_mem

[xkevio]

No description provided.

[rustbot]

Some changes occurred in library/core/src/intrinsics/gpu.rs

cc @Flakebi

Some changes occurred to the intrinsics. Make sure the CTFE / Miri interpreter
gets adapted for the changes, if necessary.

cc @rust-lang/miri, @RalfJung, @oli-obk, @lcnr

[rustbot]

r? @clarfonthey

rustbot has assigned @clarfonthey.
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

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: libs
• libs expanded to 12 candidates
• Random selection from 6 candidates

[RalfJung]

#135516 does not mention this feature either?

View changes since the review

[xkevio]

It is the one mentioned by @Flakebi as the Tracking Issue in #146181 which implements this feature, so I figured it might be more appropriate.

[Flakebi]

#135516 is the only shared memory issue I know of (thanks for fixing the typo!). I can create a new one if that’s more appropriate

[RalfJung]

We typically have one tracking issue per feature. The issue should explicitly say which feature it is for. The nvptx situation seems to be quite messy with #135516 and #111199 both saying they are tracking stdarch_nvptx. Neither of them seems to be for gpu_launch_sized_workgroup_mem. Maybe this intrinsic should be moved to stdarch_nvptx but that sounds odd as it the intrinsic sounds more cross-platform than that.

Someone who knows all the relevant features here should go through this all and clean it up so that each feature has exactly one tracking issue that reflects what the feature is about and what the current status us.

[xkevio]

Yeah, it does seem like the better solution is to create a more general tracking issue, maybe for intrinsics::gpu in general? The two intrinsics in there are cross-platform after all, and it will only grow over time I assume.

[RalfJung]

Taking a step back -- intrinsics are not meant to be stabilized ever. Instead we add public functions in a proper place that invoke the intrinsics, and only those functions ever get stabilized. Why does this one have a feature gate?

[clarfonthey]

@rustbot author

Yeah, intrinsics generally don't have stability attributes; we only stabilize other functions that call them. Will wait for the discussion to decide what to do with this.

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.