Add 16-bit and 32-bit FFI types #6060
Conversation
|
agreed, void isn't; joking but serious |
pchiusano
left a comment
There was a problem hiding this comment.
Nice!!
Though I can't tell if CI has actually passed yet.
Not yet |
|
I'm not sure what's going on with the runtime tests. Doesn't seem related to FFI but still? |
|
The tests are failing because it's getting the wrong kind of error. It's expecting I don't understand the test well enough to know how this could possibly be related to what I changed, though. But I reran one of these in another PR and it didn't seem to fail. Maybe @stew or @ceedubs has a better idea? Edit: actually, if I run it locally on this and a different branch, both seem to fail. So I think it's unrelated to this PR. Is it just some transient problem with example.com? |
|
I'm not sure what's happening now. It can't lock the codebase? |
|
@dolio I cleared the cache entries, which no one should ever have to do, and then re-ran it, and that thankfully worked 
|
|
Somehow it cached a locked file? I don't get it. |
This adds support for 16-bit integers and 32-bit integers and floats to the DLL based FFI. The unison analogues used are the 64-bit versions of each, and they are down/up converted at the boundary with C.
I added test cases for each to the transcript, so we should be testing all the FFI types now (except void, which isn't