building v17 on MacOS
-
- Posts: 4
- Joined: Thu 19 Nov, 2020 5:18 pm
building v17 on MacOS
I'm wondering if anyone here has successfully managed to compile v17 for macOS using the previously posted directions at https://pcem-emulator.co.uk/phpBB3/view ... 863#p13863.
I gave it a shot but ran into a lot of new errors regarding wx, so I'm still hanging out on v16.
I gave it a shot but ran into a lot of new errors regarding wx, so I'm still hanging out on v16.
Last edited by StLouisCPhT on Sat 16 Jan, 2021 11:28 pm, edited 1 time in total.
Re: v17 on MacOS
I've managed to compile v17 under Mac OS catalina and the latest XCode by adding
to wx-threads.c, and
to slirp/slirp.h
Oddly, now I have to add the includes event to compile v16, while the last time I worked on the code (before the v17 feature freeze) they where not needed.
Code: Select all
#include <sys/time.h>
Code: Select all
#include <stdint.h>
Oddly, now I have to add the includes event to compile v16, while the last time I worked on the code (before the v17 feature freeze) they where not needed.
Re: v17 on MacOS
I was able to compile v17 on Mac OS with no additional changes from the original instructions (I'm on OSX 10.14.6) but I've noticed that some of my machines now run dog slow. If I copy the pcem executable v16 back over everything runs normal again. It seems to affect the PMMX class machines.
-
- Posts: 4
- Joined: Thu 19 Nov, 2020 5:18 pm
Re: v17 on MacOS
Your addition to wx-thread.c got rid of the time error.davide78 wrote: ↑Mon 07 Dec, 2020 9:40 amI've managed to compile v17 under Mac OS catalina and the latest XCode by addingto wx-threads.c, andCode: Select all
#include <sys/time.h>
to slirp/slirp.hCode: Select all
#include <stdint.h>
Oddly, now I have to add the includes event to compile v16, while the last time I worked on the code (before the v17 feature freeze) they where not needed.
Adding #include <unistd.h> to slirp/slirp.h dealt with clang reporting implicit declaration of function close(s) errors in the following spots:
slirp/debug.c
slirp/misc.c
slirp/socket.c
slirp/top_subr.c
slirp/tftp.c
slirp/udp.c
Just a heads up: #include <stdint.h> exists at line 80 in slirp/slirp.h, so you shouldn't need to add it yourself.
Now to figure out how to get my finished build of v17 to stop having a fit about where wx is at over on my ARM MacBook Pro.
Re: v17 on MacOS
You are right. Indeed, in the actual code I added #include <unistd.h>. I copy-pasted the wrong line of code in my previous message.StLouisCPhT wrote: ↑Wed 09 Dec, 2020 12:43 amAdding #include <unistd.h> to slirp/slirp.h dealt with clang reporting implicit declaration of function close(s) errors in the following spots:
.....
Just a heads up: #include <stdint.h> exists at line 80 in slirp/slirp.h, so you shouldn't need to add it yourself.
-
- Posts: 4
- Joined: Thu 19 Nov, 2020 5:18 pm
Re: v17 on MacOS
The reason for some of your machines are behaving so freaking slow is that the Dynamic Recompiler is completely broken on macOS with (pcem) v17, so every machine that relies only on Dynarec (like Pentium-class CPUs) are unusablely slow.TonyM wrote: ↑Mon 07 Dec, 2020 5:11 pmI was able to compile v17 on Mac OS with no additional changes from the original instructions (I'm on OSX 10.14.6) but I've noticed that some of my machines now run dog slow. If I copy the pcem executable v16 back over everything runs normal again. It seems to affect the PMMX class machines.
All that is left for us is downgrade to v16 again and hope that someday someone could fix the dynarec for macOS (which I would not expect much since there is such a low interest in porting properly Pcem for macOS).
-
- Posts: 4
- Joined: Thu 19 Nov, 2020 5:18 pm
Re: v17 on MacOS
That comment is not exactly going to encourage Sarah or anyone else to help out.
And making a post that is not even relevant to the post topic (building v17 on macOS) is bad form also.
Last edited by StLouisCPhT on Sat 16 Jan, 2021 11:26 pm, edited 1 time in total.
Re: v17 on MacOS
I wasn't trying to motivate anyone. It's up to her to code whatever she wants to. I just wanted to give a technical approach to this topic since no one ever mentioned that the problem was with the Dynamic Recompiller before.
So yes, it was quite relevant from the technical point of view.
It's your comment just to criticize me that didn't added anything relevant to this discussion.
Kthxbye.
So yes, it was quite relevant from the technical point of view.
It's your comment just to criticize me that didn't added anything relevant to this discussion.
Kthxbye.