Followup review feedback on e2e codegen example

[omarahmed1111]

This PR is a follow up PR to address the rest of review feedback on 610. It contains:

• readme file to instruct how to run the codegen example.
• filter out adapters that doesn't support spirv kernels from codegen example.
• print the input array to the spv kernel in the codegen example.

Fixes: 859

[omarahmed1111]

is this for page alignment?
getpagesize() ?

It would make since that this is related to gpus page size, but not sure how to make sure of that. I think it will need to be generalized if it will run on anything other than L0.

[omarahmed1111]

Now that we have GPU-equipped runners, it should be possible to create a workflow where this example is built and run as part of tests. Otherwise, it will be easy to accidentally break it.

I would think that would still be blocked by this issue, so it will be hard to test that while failure, I think that could be added as a seperate PR with a seperate issue when the L0 pipeline is fixed.

[pbalcer]

is this for page alignment?
getpagesize() ?

It would make since that this is related to gpus page size, but not sure how to make sure of that. I think it will need to be generalized if it will run on anything other than L0.

I guess 4k is a safe enough value... I'd still do a define/const variable instead of having a magic number though.

[pbalcer]

Now that we have GPU-equipped runners, it should be possible to create a workflow where this example is built and run as part of tests. Otherwise, it will be easy to accidentally break it.

I would think that would still be blocked by this issue, so it will be hard to test that while failure, I think that could be added as a seperate PR with a seperate issue when the L0 pipeline is fixed.

The L0 workflow is not blocked, that test is simply not ran in CI (plus there's a timeout). But I agree that it might make sense to do this in another PR.

[omarahmed1111]

Now that we have GPU-equipped runners, it should be possible to create a workflow where this example is built and run as part of tests. Otherwise, it will be easy to accidentally break it.

I would think that would still be blocked by this issue, so it will be hard to test that while failure, I think that could be added as a seperate PR with a seperate issue when the L0 pipeline is fixed.

The L0 workflow is not blocked, that test is simply not ran in CI (plus there's a timeout). But I agree that it might make sense to do this in another PR.

I created this issue to track this. I was trying to attach it to this pr but it seems that it will not work directly and might be controversial in some parts. so, will submit it to a new pr.

[lukaszstolarczuk]

👍