gollvm: meaningless "unsupported llvm::CallingConv::ID"

38 views
Skip to first unread message

Ivan Serdyuk

unread,
Jan 14, 2021, 8:38:15 AM1/14/21
to golang-nuts
Hi.

I caught the following error:
.

Among other things: people has nothing to do with an ID. There is a specific list of options:
.
But providing a human friendly name with an ID could have some sense.

Another problem is that this is an intersection with the class, dedicated to a specific struct alignment. I have doubts that all (at least - user mode, not kernel mode ) calling conventions would assume all possible options, for aligning structs/unions. Also there are OS specific issues - so it more about OS-convention-alignment triple.

I would propose some re-design, to offer a purpose specific output, depending on developer's goals. So the output would help to navigate against the next step.

Ivan

Ivan

Than McIntosh

unread,
Jan 14, 2021, 9:03:01 AM1/14/21
to Ivan Serdyuk, golang-nuts
On Thu, Jan 14, 2021 at 8:38 AM Ivan Serdyuk <local.tou...@gmail.com> wrote:
Would it help to share the actual error message and compiler invocation?
Among other things: people has nothing to do with an ID. There is a specific list of options:
.
But providing a human friendly name with an ID could have some sense.

Not sure what you are saying here. 

The line number you cite above indicates that gollvm is being asked to use a calling convention that is recognized and/or supported by clang/LLVM, but not supported by gollvm.  

Another problem is that this is an intersection with the class, dedicated to a specific struct alignment. I have doubts that all (at least - user mode, not kernel mode ) calling conventions would assume all possible options, for aligning structs/unions. Also there are OS specific issues - so it more about OS-convention-alignment triple.

I would propose some re-design, to offer a purpose specific output, depending on developer's goals. So the output would help to navigate against the next step.

?

Again, not sure what you are getting at.

Than

Reply all
Reply to author
Forward
0 new messages