Emulator

I can't reproduce with my copy of the game (fe441d77d5cb9a7f01de954e84e6ca49, revision 0 - yours is revision 2). For mine, the game more or less immediately closes when stopping emulation (without any kind of fadeout, which seems to be normal for launch games - I wouldn't be surprised if revision 2 behaves differently with regards to this).

Note that with Wii games, stopping emulation simply tells the emulated game that the power button has been pressed, and then waits for the game to do its stuff (e.g. finish saving). In some cases, the game isn't listening for the signal (mostly if you press the power button during early startup). You can click the stop emulation button a second time to force emulation to stop (without asking the game nicely).

Can you post a log file with "IOS - State Transition Manager" enabled and the log level set to info? That might contain additional data that helps here.

So I am not sure if I did what you said correctly but here is the output I got.

45:01:437 Core/IOS/STM/STM.cpp:46 I[IOS_STM]: /dev/stm/immediate - IOCtl:
45:01:437 Core/IOS/STM/STM.cpp:47 I[IOS_STM]: IOCTL_STM_VIDIMMING

Even after waiting a few mins and even after forcing shut down. Nothing changed.

I ran the same info with another game and it showed the shutdown command like I think you were looking for.

Yes, that's the information I was looking for, thanks. Unfortunately the lack of any shutdown command means that I don't have any ideas for further steps to take (though a developer with a copy of the revision 2 of Super Paper Mario could investigate further).

I can reproduce the issue - the game freezes in the middle shutdown for me.

Don't know if this helps but it looks like it isn't acknowledging the shutdown command at all and the other processes continue. I toggled on all logging and found this.

10:25:236 m_stop_internal:2675 N[Audio]: Cubeb stream (0x149e3c100) stopped successfully. (I hit the red X to close the window and it said stop. Then after hitting yes to shutdown it then produced the next line. Each stop line is me hitting the X. When asked to force shutdown I hit no.)
10:27:346 m_stop_internal:2647 N[Audio]: Cubeb stream (0x149e3c100) started successfully.
10:36:009 m_stop_internal:2675 N[Audio]: Cubeb stream (0x149e3c100) stopped successfully.
10:41:555 m_stop_internal:2647 N[Audio]: Cubeb stream (0x149e3c100) started successfully.
10:46:531 m_stop_internal:2675 N[Audio]: Cubeb stream (0x149e3c100) stopped successfully.
10:51:480 m_stop_internal:2647 N[Audio]: Cubeb stream (0x149e3c100) started successfully.

Using LLE Audio fixes the issue. Obviously not a great option, but at least the more accurate option works.

Thanks for the workaround. I will also report that leaving it on HLE and changing the backend setting from Cubeb to No Audio Output still causes the issue so it seems specific to the HLE.

I can confirm this issue multiple times because I'm currently playing this game (rev 2). I didn't realize it is a bug.

JMC provided some screenshots of the threads widget (with symbols generated from the signature database): https://usercontent.irccloud-cdn.com/file/KqbWN3hl/image.png https://usercontent.irccloud-cdn.com/file/ioiWWsPM/image.png https://usercontent.irccloud-cdn.com/file/DGf7LYcY/image.png https://usercontent.irccloud-cdn.com/file/x4TG7dg3/image.png https://usercontent.irccloud-cdn.com/file/z9350saK/image.png - the hang in __AIDHandler is apparently related to DSP HLE timings.

spm-practice-codes includes a patch for this: inserting a blr right at the start of spm::spmario_snd::spsndExit on versions other than US rev 0, Japan rev 0, or Japan rev 1. (It also provides a list of addresses for that function in all revisions). We could theoretically include this patch, though it'd be better to just fix the timing issue (if we can figure it out).

This also impacts Sakura Wars: So Long My Love (Wii)