1 |
On Thu, Apr 28, 2016 at 1:37 PM, Leno Hou <lenohou@×××××.com> wrote: |
2 |
> There is a bootstrap tarball for ppc64le now, and we have bi-arch |
3 |
> support for ppc64 platform. It's means that we can build go-1.6.2 |
4 |
> without gccgo |
5 |
> |
6 |
> Signed-off-by: Leno Hou <lenohou@×××××.com> |
7 |
> --- |
8 |
> dev-lang/go/go-1.6.2.ebuild | 5 +++++ |
9 |
> 1 file changed, 5 insertions(+) |
10 |
> |
11 |
> diff --git a/dev-lang/go/go-1.6.2.ebuild b/dev-lang/go/go-1.6.2.ebuild |
12 |
> index 57ca2c7..ba61397 100644 |
13 |
> --- a/dev-lang/go/go-1.6.2.ebuild |
14 |
> +++ b/dev-lang/go/go-1.6.2.ebuild |
15 |
> @@ -88,6 +88,11 @@ go_arch() |
16 |
> case "${portage_arch}" in |
17 |
> x86) echo 386;; |
18 |
> x64-*) echo amd64;; |
19 |
> + ppc64) if [[ "$(tc-endian $@)" == "big" ]]; then |
20 |
> + echo ppc64 |
21 |
> + else |
22 |
> + echo ppc64le |
23 |
> + fi;; |
24 |
> *) echo "${portage_arch}";; |
25 |
> esac |
26 |
> } |
27 |
> -- |
28 |
> 2.7.3 |
29 |
> |
30 |
> |
31 |
|
32 |
Sometimes, a nested 'case' construct may prove a bit clearer. Just a |
33 |
suggestion. Not being critical of the original implementation in any |
34 |
way. |
35 |
... |
36 |
ppc64) |
37 |
case $(tc-endian $@) in |
38 |
big) |
39 |
echo ppc64 |
40 |
;; |
41 |
little) |
42 |
echo ppc64le |
43 |
esac |
44 |
... |