Re: [hatari-devel] Mac / PortMidi issues (was: Fixing some STE sound issue and doing 2.4.1 release)

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On 3.8.2022 23.37, Nicolas Pomarède wrote:
> Le 02/08/2022 à 16:01, Eero Tamminen a écrit :
> > On 2.8.2022 15.08, Thomas Huth wrote:
> > > On 02/08/2022 14.02, Nicolas Pomarède wrote:
> > > > now that all known memory leaks have been fixed (expect the SDL 
> > > > ones), are there any remaining issues or changes to be done before 
> > > > releasing 2.4.1 ?
> > >
> > > At least I don't have anything left.
> >
> > I don't have anything left either.
>
> release process in progress, files have been tagged to 2.4.1 in source 
> tree.
>
> I'm waiting for the windows / macOS version to be compiled to make the 
> announcement later.

Some Mac issues are reported here:
https://www.atari-forum.com/viewtopic.php?p=436440#p436440

Both for v2.4.0 release (crash in devices SDL GUI dialog [1]), and for 
the latest snapshot (does not start).

You might want to wait until at least latter is resolved before 
announcing new release.


	- Eero

[1] I hadn't run ASAN with PortMidi enabled before, but that does not 
report any access for me with latest Hatari git.

There are two minor leaks though:
----------------------------------------
Direct leak of 77 byte(s) in 1 object(s) allocated from:
    #0 0x7f8d603ece8f in __interceptor_malloc 
.../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:145
    #1 0x7f8d5fc59842 in find_default_device 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x4842)
    #2 0x7f8d5fc596d5 in pm_init 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x46d5)
    #3 0x7f8d5fc5a559 in Pm_Initialize 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x5559)
    #4 0x7f8d5fc5a578 in Pm_CountDevices 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x5578)
    #5 0x5586bd2f18e0 in Midi_Host_GetPortName src/midi.c:442
    #6 0x5586bf5d5bf5 in Dialog_DeviceDlg src/gui-sdl/dlgDevice.c:138

Direct leak of 76 byte(s) in 1 object(s) allocated from:
    #0 0x7f8d603ece8f in __interceptor_malloc 
.../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:145
    #1 0x7f8d5fc59842 in find_default_device 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x4842)
    #2 0x7f8d5fc596bc in pm_init 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x46bc)
    #3 0x7f8d5fc5a559 in Pm_Initialize 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x5559)
    #4 0x7f8d5fc5a578 in Pm_CountDevices 
(/lib/x86_64-linux-gnu/libportmidi.so.0+0x5578)
    #5 0x5586bd2f18e0 in Midi_Host_GetPortName src/midi.c:442
    #6 0x5586bf5d5bf5 in Dialog_DeviceDlg src/gui-sdl/dlgDevice.c:138
----------------------------------------

(This was first thing requiring ASAN_OPTIONS=fast_unwind_on_malloc=0 for 
proper backtraces.  Maybe my portmidi lib is missing frame-pointers 
and/or unwind info section.)