Skip to content

fix(gtk): bump go-gtk/gtk4 to v0.2.2 (native GTK on loong64/riscv64/ppc64le) - #92

Closed
tannevaled wants to merge 1 commit into
mainfrom
fix/gtk-v0.2.2
Closed

fix(gtk): bump go-gtk/gtk4 to v0.2.2 (native GTK on loong64/riscv64/ppc64le)#92
tannevaled wants to merge 1 commit into
mainfrom
fix/gtk-v0.2.2

Conversation

@tannevaled

Copy link
Copy Markdown
Contributor

go-gtk/gtk4 v0.2.2 narrows the purego guard to only s390x — loong64/riscv64/ppc64le get the real GTK backend instead of the stub (purego supports them). The qemu lanes there dlopen no libgtk and skip cleanly; s390x keeps the stub. All GOOS/GOARCH build clean.

…ppc64le too

go-gtk/gtk4 v0.2.2 narrows the purego guard to only s390x (big-endian, which
purego lacks). loong64/riscv64/ppc64le now get the real GTK backend instead of the
stub; the qemu test lanes there dlopen no libgtk and skip cleanly. s390x keeps the
stub until purego gains the big-endian FFI trampoline.
@tannevaled

Copy link
Copy Markdown
Contributor Author

Abandoning: purego runs under qemu-arm64 but not qemu-loong64/riscv64/ppc64le, so the real binding there breaks the qemu CI lanes. Those arches are little-endian (no bug-exhibition value; only s390x is big-endian, and it's stubbed either way). The window stays on go-gtk v0.2.1's amd64||arm64 guard (v0.61.1) — native GTK where it matters and where qemu can test it, drawn fallback elsewhere.

@tannevaled tannevaled closed this Aug 31, 2026
@tannevaled
tannevaled deleted the fix/gtk-v0.2.2 branch August 31, 2026 16:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant