ominitay
ominitay
Disabled the failing behaviour tests caused by the bug in the x86 backend. My rationale here is that it's not within the scope of this PR to work around, as...
Possibly related to #12240
Ooooop I didn't spot this Hopefully nothing??
So running a `git diff`, there are a couple slight things missing from theirs that are included in mine, but they should be functionally the same.
> Is this the correct place to complain about `os.argv`? (It's defined as {} on Windows which means code will compile but almost always panic as argv.len is generally considered...
Another? I don't think that `start.zig` performs any heap allocation as of present. And I don't believe that we should change that. It also guarantees a memory leak if a...
> There's TLS on Linux, which does it through mmap. Didn't spot that. TLS avoids using `mmap` if it can fit things in a buffer, but is also necessary to...
Additionally, it makes things simpler for library authors etc than saying *maybe* `argv` will compile error on Windows, or *maybe* it won't.
> maybe it could be opt-in This would defeat the point of such a change.
I wholly agree with the third option, following the Zig Zen: `Only one obvious way to do things.`