Re: Bug#960674: golang-go: "fatal error: gc_trigger underflow" on mipsel

Shengjing Zhu <[email protected]> 于2020年6月25日周四 上午1:59写道：
>
> Hi syq,
>
> On Thu, May 28, 2020 at 9:04 AM YunQiang Su <[email protected]> wrote:
> [...]
> > you are right. the current golang still output FP32 object...
> > So, we think that it is buggy.
> >
> > Since Loongson CPU has some strange behaviour, it even can work...
> > Let's try to patch golang to support FPXX or FP64.
> >
> > https://github.com/golang/go/issues/39289
> >
>
> Do these two patches[1] by you can fix the problem we currently have?

Yes.

> It's annoying currently since it blocks testing migration of all Go
> packages.
> Even upstream hasn't merged these, we can land these in Debian.
>
> [1] https://go-review.googlesource.com/c/go/+/237058/

It is approved by upstream. It will be safe to use it for Debian.

>  https://go-review.googlesource.com/c/go/+/239217/

It hasn't reviewed by upstream. and it is useful for only one case:
   build a pure go code to ELF object
   link it with a standard C shared object (libc etc)
If we don't have this case, you can just ignore it.

>
> If they are ready for Debian, let's upload a mipsel version and then binnmu it.
>
> --
> Shengjing Zhu