mirror of
https://codeberg.org/ziglang/zig
synced 2026-11-18 07:05:05 +00:00
Native detection of the libc version now only happens if the ABI component in the triple is native (= omitted in text form), rather than the OS component. So the ABI is either fully native or fully specified with no weird in-between. I think this much more closely matches people's mental model for this stuff; Matthew and I struggled to make sense of the current semantics for almost an hour, and ultimately agreed that it really shouldn't work the way it did prior to this patch. This is technically a breaking change because `native-native-gnu` no longer means "use glibc with the detected native glibc version", and this concept is no longer expressible. I don't consider this a great loss; I couldn't find any usage in the wild that would be broken by this, and I would argue it was more likely to be an unhelpful surprise. On the other hand, this patch makes `native-native-gnueabihf` on e.g. an `x86_64-linux-gnu` system use the default glibc version rather than the native one - which was a behavior I found bizarre. I think the ABI detection code is now also much easier to reason about. closes https://codeberg.org/ziglang/zig/issues/36863 |
History
|
|
|---|---|---|
| .. | ||
| aarch64.zig | ||
| alpha.zig | ||
| amdgcn.zig | ||
| arc.zig | ||
| arm.zig | ||
| avr.zig | ||
| bpf.zig | ||
| csky.zig | ||
| generic.zig | ||
| hexagon.zig | ||
| hppa.zig | ||
| kvx.zig | ||
| lanai.zig | ||
| loongarch.zig | ||
| m68k.zig | ||
| mips.zig | ||
| msp430.zig | ||
| nvptx.zig | ||
| powerpc.zig | ||
| propeller.zig | ||
| Query.zig | ||
| riscv.zig | ||
| s390x.zig | ||
| sparc.zig | ||
| spirv.zig | ||
| ve.zig | ||
| wasm.zig | ||
| x86.zig | ||
| xcore.zig | ||
| xtensa.zig | ||