dmed (David Medina)
dmed (David Medina)
Right now, `@inner` loops need to share the same number of iterations. However, `@exclusive` arrays are hard-coded to `256` entries which creates hard-to-debug bugs I agree there should be a...
@pdhahn Make sure to use ``` ` ``` to escape `@attributes` to prevent emailing random people
np, I was accidentally emailing `@dim` a lot and he kindly replied with that info
The `@inner` loop size restriction is due to GPU blocks/work groups needing to be the same size :(
Yeah, based on the number of iterations and like the docs say > The concept of exclusive memory is similar to thread-local storage, where a single variable actually has one...
I think your first interpretation was right, I meant the concept was similar - TLS: 1 thread - 1 value - `@exclusive`: 1 iteration - 1 value > it would...
@jlchan sorry about that! Maybe we should increase the number as a temporary fix
Hi @pdhahn ! Commercial deployment of OKL code seems like an interesting feature to add. To be honest it might not be super high priority at the moment due to...
Just pushed [f4fea623f09f12249ab2c783a5be6730f5497681] which adds a `hash()` method to the `occa::kernel` object. The way to get the binary would be ```cpp occa::kernel myKernel = ...; std::string hash = myKernel.hash().getString(); ```...
Structs of structs with basic types (like int/double/nested structs in #141) is probably the next feature to be added, so stay tuned :) I don't think we can handle pointers...