Re: [Sbcl-devel] [Sbcl-commits] master: x86-64, struct-by-value: correctly pass small structs on the stack

[Stas B. <sta...@gm...>]

I just send c-call.lisp to various free LLMs and say "find bugs here".
And also make them generate test cases.
I didn't include any specific test because they won't find new problems.
A test that performs an exhaustive combination of different foreign
types is needed.
But I'm not eager to make that yet.

One of the bugs was hard to trigger (has been there for 10 years), it
was writing reading to D2 from the stack instead of D0, but the
writing part was using STR D0, [stack], so most of the time it worked
without an issue.
Hard to write a test for something like that.

On Thu, Apr 23, 2026 at 10:36 PM Christophe Rhodes via Sbcl-devel
<sbc...@li...> wrote:
>
> stassats via Sbcl-commits <sbc...@li...> writes:
>
> > The branch "master" has been updated in SBCL:
> >        via  627064c1dca8cda8b982873e9bf1b784da245b46 (commit)
> >       from  467666387e8a978e6580277e241d4011b04d629c (commit)
> >
> > - Log -----------------------------------------------------------------
> > commit 627064c1dca8cda8b982873e9bf1b784da245b46
> > Author: Stas Boukarev <sta...@gm...>
> > Date:   Thu Apr 23 20:15:52 2026 +0300
> >
> >     x86-64, struct-by-value: correctly pass small structs on the stack
>
> Clearly our existing tests were insufficient... do you have distillable
> test cases for these changes?  How are you finding the problems /
> validating the fixes?
>
> Christophe
>
>
> _______________________________________________
> Sbcl-devel mailing list
> Sbc...@li...
> https://lists.sourceforge.net/lists/listinfo/sbcl-devel