[zfs] Fix LZ4 function definitions to not trip up the compile

[adrian]

Summary

This is today (Nov 16 2020) head amd64; llvm tripped up in compiling
userland by noting that lz4_init/lz4_fini were lacking in being previously
defined.

So, just remove the extern and add it to lz4. So far so good.

Test Plan

• compiled userland/kernel on amd64

[brooks]

How did this work for anyone? I tripped on it in CheriBSD on aarch64 just now (this is with llvm ~11.0.0).

[dim]

In D27240#608620, @brooks wrote:

How did this work for anyone? I tripped on it in CheriBSD on aarch64 just now (this is with llvm ~11.0.0).

What was the exact error message? I never got that, and it always compiled just fine for me...

[dim]

Apparently for me, _FAKE_KERNEL just never gets defined, so that part of the code never gets compiled. I have no idea where that define comes from. In the preprocessed lz4.i I simply do not see the offending code fragment at all.

[imp]

In D27240#608670, @dim wrote:

Apparently for me, _FAKE_KERNEL just never gets defined, so that part of the code never gets compiled. I have no idea where that define comes from. In the preprocessed lz4.i I simply do not see the offending code fragment at all.

I'm guessing it's a bit like our _STANDALONE environment, but that's less well defined in the upstream code.