| 1 |
On Fri, 2009-11-20 at 18:34 +0100, Fabian Groffen wrote: |
| 2 |
> On 06-10-2009 08:34:14 +0100, Alan Hourihane wrote: |
| 3 |
> > > yeah, but back then it didn't rm -Rf internal/libffi |
| 4 |
> > |
| 5 |
> > The Modules/_ctypes/libffi is the same as the external version. So no, |
| 6 |
> > it won't compile. |
| 7 |
> |
| 8 |
> We've been asked to file bug about these issues with upstream libffi, so |
| 9 |
> could you please paste a bunch of compilation errors somewhere? |
| 10 |
|
| 11 |
libffi supports m68k, but uses specific cacheflush syscalls that |
| 12 |
FreeMiNT doesn't support. So it needs a specific FreeMiNT patch. |
| 13 |
|
| 14 |
Additionally, it seems that 68020 or higher is required, whereas some of |
| 15 |
our targets in FreeMiNT are 68000 processors. So some of the m68k |
| 16 |
assembly code needs re-writing for this. |
| 17 |
|
| 18 |
Alan. |
Archives