Thanks for this diff and the interest in the port. On Fri, May 29, 2020 at 03:15:11PM +0200, Omar Polo wrote: [...] > On my machine (amdgpu) it leaves a core file around but otherwise is > working; on my friend machine (inteldrm) it doesn't core dumps. We
Can you provide some details about the core dump? Does Godot crash on amdgpu? Can you share a backtrace with gdb(egdb) from ports? I don't understand if this is just like the already known bugs with amdgpu at this point or something new. > working; on my friend machine (inteldrm) it doesn't core dumps. We > tested everything but the networking stuff AFAIK. As your patch includes disabling ssl in favor of mbedtls, I would think that some networking should be tested. > I didn't have the time (yet) to fully debug this issue. > > Do I have to reset the REVISION back to 0 since this changes the port > version? Yes, REVISION is only used if changes happen in the package while the version doesn't change. It effectively adds pX to the package. See packages-specs(7) and bsd.port.mk(5). So just remove the REVISION line when updating the port's version. [...] > - pulseaudio is still disabled: I have a WIP patch (not included) to > bring back audio support using libao. I plan to submit it soon > after this update I would be _very_ interested in this after the update. I've been trying to come up with an sndio backend myself, but haven't got it to work so far. [...] > - in 3.1 they replaced openssl with mbedtls, hence the removal of ssl > from WANTLIB Must be a bundled mbedtls, as nothing is added to WANTLIB and/or LIB_DEPENDS. Might be better to find a way to use our ports version... [...] I haven't tested this yet. I am comparing it with the diff that I have been working on and will get back to you. Your diff is much longer than what I had (537 vs 199 lines). I'm gonna see what we need. Port updates are usually preferred as inline diff. Just make sure your mail client doesn't mangle whitespace.