[BUG]: cuda::std::extents::extent does not work in constexpr context if there is at least one dynamic_extent present

[pciolkosz]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this bug and that I agree to the Code of Conduct

Type of Bug

Compile-time Error

Component

libcu++

Describe the bug

It's not possible to use extents::extent in for eg. static_assert even when statically known extent is queried if extents contain at least one dynamic_extent. If all extents are statically known, then extent works in constexpr contexts.

How to Reproduce

https://godbolt.org/#z:OYLghAFBqd5QCxAYwPYBMCmBRdBLAF1QCcAaPECAMzwBtMA7AQwFtMQByARg9KtQYEAysib0QXACx8BBAKoBnTAAUAHpwAMvAFYTStJg1DIAruiakl9ZATwDKjdAGFUtEywYgATKUcAZPAZMADl3ACNMYhAADlIAB1QFQjsGFzcPPQSk2wEAoNCWCKiuS0xrHIYhAiZiAjT3Tx8rTBsUqpqCPJDwyJjLatr6jJKFAc7A7sLergBKS1QTYmR2DgBSLwBmQOQ3LABqVY2nU3MAelH0U5Z0BTjDQ%2BxVjQBBJ%2BfAgj2WJkCIGYOAOwAITeezBexOTBAIAu0MwqgIjAICkOTg%2BpD2JQhZihMII6Gh6AAnswWHhkAB9eGIwQYjYPPbwiBeGaHEEvcF7UZMWyUpgKJS1CDwgB01KREA0/0OABFZZjWRt2c9OZDobCQOLBCijujMRifHt6RtsIzVF42aDwdzeRT%2BYKCMLzWKERKpQcNnLPQrLS9VgCZRw5rROABWXieDhaUioThOOQylVOdYgrxArkLJaYA6bHikAiaINzADWIFDAE4ReWAGxecteaLVjYA2sAyQbUP6TiSCOFmOcXgKEAafOFuZwWAwRAoVAsOJ0SLkShoOcLqIMABuyGQXC8XhMJSwG/JmAAanhMAB3ADycUYfZotERxCHEDCfbCgRqRI/X%2BIROvMJtBaAtuF4Fc2EEa8GFoH8o14LAwhMYAnDEWghzA0gsG%2BIxxHgrC8GIEC8A3TAMOjeEWhMRE%2Bw%2BMo%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

Expected behavior

Both static asserts from the godbolt link should compile correctly, presence of dynamic_extent should not affect the ability to query statically known extents at compile time.

Reproduction link

https://godbolt.org/#z:OYLghAFBqd5QCxAYwPYBMCmBRdBLAF1QCcAaPECAMzwBtMA7AQwFtMQByARg9KtQYEAysib0QXACx8BBAKoBnTAAUAHpwAMvAFYTStJg1DIAruiakl9ZATwDKjdAGFUtEywYgATKUcAZPAZMADl3ACNMYhAADlIAB1QFQjsGFzcPPQSk2wEAoNCWCKiuS0xrHIYhAiZiAjT3Tx8rTBsUqpqCPJDwyJjLatr6jJKFAc7A7sLergBKS1QTYmR2DgBSLwBmQOQ3LABqVY2nU3MAelH0U5Z0BTjDQ%2BxVjQBBJ%2BfAgj2WJkCIGYOAOwAITeezBexOTBAIAu0MwqgIjAICkOTg%2BpD2JQhZihMII6Gh6AAnswWHhkAB9eGIwQYjYPPbwiBeGaHEEvcF7UZMWyUpgKJS1CDwgB01KREA0/0OABFZZjWRt2c9OZDobCQOLBCijujMRifHt6RtsIzVF42aDwdzeRT%2BYKCMLzWKERKpQcNnLPQrLS9VgCZRw5rROABWXieDhaUioThOOQylVOdYgrxArkLJaYA6bHikAiaINzADWIFDAE4ReWAGxecteaLVjYA2sAyQbUP6TiSCOFmOcXgKEAafOFuZwWAwRAoVAsOJ0SLkShoOcLqIMABuyGQXC8XhMJSwG/JmAAanhMAB3ADycUYfZotERxCHEDCfbCgRqRI/X%2BIROvMJtBaAtuF4Fc2EEa8GFoH8o14LAwhMYAnDEWghzA0gsG%2BIxxHgrC8GIEC8A3TAMOjeEWhMRE%2Bw%2BMo%2B1oPAwmIb8XCwPsCGIPAWE4Hg5ioAxgAUc8r1ve9MP4QQRDEdgpBkQRFBUdR8N0EoDCMFAcX0Jih0gOZUDiCoMN4VBSOILisF0v5SnKFIHAYZxXAaPR/AmAoikyRJkgEIZPBKLJvIYLp3OmGziIEdpBic4YwtaCKxmCnpin6DpfL0blakSqZijmBRM2WCRgzDXt8NjDg9mCU8nCcTEvBFOquD2CBcEIEgcw2WZeFArQZhLMsRxDDge1IHiNg2EUxsmqbpurUhI2jMrB2HUd4PHKcICQBYCDiail2FfAiAsvRJOEURxDkk7FLUPtVNIS8WLiMcuw4cM5r7Mrr2onbPlQKgKqqmrd3qkVGogFxV3oYh2s6laermBBMCYLAomswbhtG8bpqxybZvmkyB0sZbuqDUgJ2nI8Tz2ld50h4JWBWSrqtq4G8wpgqOF4TADpIPACRKE7pPO6RLqUa6VPSspws8YUHLSkpXPyJLPOyFI5fiLyKiyjyRkluLKjGNXmj1yLxkV7L0oN6K/JSzK3KV2Z8yIzBeeHIqXpKhbOBlTBj2WPYRMvSJ/qZoGGqalrDuhjFwZpoP1g6mYurHN30ZAMaJuxrG3tKgmhxHYnetIUtJA0EUNGbDQAS4DZa2bLxQ2rTtBo2D38Y52Gi2erxW/7duC7mMyknsSQgA

Operating System

N/A

nvidia-smi output

N/A

NVCC version

12.2.1

[github-actions]

Hi @pciolkosz!

Thanks for submitting this issue - the CCCL team has been notified and we'll get back to you as soon as we can!
In the mean time, feel free to add any relevant information to this issue.

[jrhemstad]

Thanks @pciolkosz, this does indeed look wrong.

@youyu3 @miscco could you take a look?

[youyu3]

Looks like a bug in our implementation. The kokkos reference implementation looks okay (https://godbolt.org/z/obWK5dYT3).

I'll take a look.

[youyu3]

Looks to me that the kokkos reference implementation works for this case because it's extents.hpp was completely rewritten (kokkos/mdspan@08a38f). I think the easiest way to fix this is to sync it up.