maddyscientist comments

Results 133 comments of


                                            maddyscientist

Implement `std::cout`

Definitely interest from another user here for this sweet feature 😄

Feature: clang CUDA (device compilation) support

For clang you probably need to enable the use of variadic functions with `-fcuda-allow-variadic-functions`

Add `cuda::std::numeric_limits<__half>`

We could certainly use this in QUDA if implemented.

This library should be usable directly by a host compiler to make compatible host objects

Adding to the motivation for this: it turns out this is a big deal for QUDA as well and prevents QUDA from adopting `cuda::std::complex`. Both for the files that only...

Cannot use `<limits>` and `<cuda/std/limits>` in the same source file

@benbarsdell this is the same issue I reported a while ago. Did you have a chance to think about how to fix this?

Question: Are cuModules shared between kernels from same program

For the constants, could this be a good use jitify's new found linking ability: declare the `__constant__` in the offline source code, e.g., in a `.cu` file and JIT compile...

Add nvrtc_cli tool

Just wondering about getting this old PR. @benbarsdell can you make the trivial changes I noted above and we can get this merged? (we can just ignore my run-time checking...

Cooperative groups support.

Thanks @csarofeen, good to have coop support. A couple of quick things: * Can you add an example of using coop_launch to the jitify_example.cpp code? * I know @benbarsdell prefers...

Add context to cache key

This looks harmless enough. Any reason not to merge this @benbarsdell?

Is --expt-extended-lambda and --expt-relaxed-constexpr not supported?

@anshumang these restrictions stem from the fact that they rely on host-device compiler interactions, and since nvrtc is a pure GPU compiler, I guess it shouldn't be surprising that they...