Bug#1012286: RFS: zig/0.9.1-1 [ITP] -- Programming language
Hi, folks.
I tried building the package on Bookworm, but one of the tests (I think it
was "std") hung using 100% of a couple CPU cores. Hanging tests are
mentioned in this Zig issue, so perhaps it's related:
"prevent confusing debugging experience when users write garbage to stdout,
interfering with the build runner / test runner protocol"
https://github.com/ziglang/zig/issues/15091
It wasn't clear at first which test was hanging, so I patched your rules
file to reveal it:
https://salsa.debian.org/zig-team/zig/-/merge_requests/4
I was able to adapt this packaging to build Zig 0.13.0 on Bookworm, after
first building llvm-toolchain-18 from its Debian Testing source. (That
rebuild was easy enough, so perhaps a case could be made for getting it into
Backports?) I ran into two failing tests in the newer Zig version, reported
here:
https://github.com/ziglang/zig/issues/21132
https://github.com/ziglang/zig/issues/21133
The first one was due to Zig's test-cli incorrectly using the XDG_CACHE_HOME
environment variable. A fix has been committed. Until the fix lands in a
release, the problem can be avoided either by not setting that variable, or
by setting it to a directory path instead of an empty value.
>I don't have any suggestions how to get around the wasm binary issue right
>now. Let me think about it for a couple of days.
This looks relevant:
"A proposal to improve trust of blobs and precompiled packages"
https://github.com/ziglang/zig/issues/19789
Reply to: