[Bug 42818] New: MIDI Time Code Quarter Frame messages not passed to ALSA

https://bugs.winehq.org/show_bug.cgi?id=42818

            Bug ID: 42818
           Summary: MIDI Time Code Quarter Frame messages not passed to
                    ALSA
           Product: Wine
           Version: 1.6
          Hardware: x86
                OS: Linux
            Status: UNCONFIRMED
          Severity: minor
          Priority: P2
         Component: winealsa.drv
          Assignee: wine-bugs(a)winehq.org
          Reporter: mdmayfield(a)yahoo.com
      Distribution: ---

MIDI Time Code (MTC) Quarter Frame messages are not correctly passed to ALSA,
causing sync to fail when the Windows application running under Wine (such as
REAPER) is the master timekeeper.

The problem is in dlls/winealsa.drv/midi.c, lines 931-945, which tries to use
snd_seq_ev_set_sysex to send these messages. Based on the discussion in
https://bugs.winehq.org/show_bug.cgi?id=26928 this must have once worked, but
apparently with more recent versions of ALSA these messages are never actually
sent via snd_seq_ev_set_sysex (it must have become more strict in its
definition of "SysEx").

Commit 53454ce9359f4ec485aa9411a0e9b14e3d1ed0fa "Fix handling of system real
time MIDI messages" by 00cpxxx appears to address a very similar issue with
some messages, but does not address Quarter Frames.

By using that approach as a template I was able to make Quarter Frame messages
work again:
https://github.com/mdmayfield/wine/commit/ab1763e7287a7062f658c40def5bf9919f...

Without this patch I could reproduce the issue on Wine 1.6, 1.8, 1.9,and 2.4.

Once I have time I will try to create a proper patch submission. Otherwise
please feel free to use or adapt the fix from the GitHub link above.

Thanks,

Matt

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 42818] MIDI Time Code Quarter Frame messages not passed to ALSA

https://bugs.winehq.org/show_bug.cgi?id=42818

--- Comment #1 from Bruno Jesus <00cpxxx(a)gmail.com> ---
Hi, thanks for the bug report. I noticed that other messages were wrong when I
made that patch but since I had no way to test the other cases I left them as
is.

Please create a helper function and use it in the 3 cases instead of
duplicating the code, you can them send it to Wine following the guide at
https://wiki.winehq.org/Submitting_Patches Otherwise just tell me and I'll make
the change.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 42818] MIDI Time Code Quarter Frame messages not passed to ALSA

https://bugs.winehq.org/show_bug.cgi?id=42818

--- Comment #2 from Matt Mayfield <mdmayfield(a)yahoo.com> ---
(In reply to Bruno Jesus from comment #1)
> Hi, thanks for the bug report. I noticed that other messages were wrong when
> I made that patch but since I had no way to test the other cases I left them
> as is.
>
> Please create a helper function and use it in the 3 cases instead of
> duplicating the code, you can them send it to Wine following the guide at
> https://wiki.winehq.org/Submitting_Patches Otherwise just tell me and I'll
> make the change.

Thanks for the comments. I tried a couple things with a helper function but did
not find a way to get it to work cleanly.

The first case 0x03, Active Sensing) uses snd_midi_event_encode_byte while the
other two use snd_midi_event_encode - but with two different buffer sizes.

Because of that it seemed like a helper function would become overly
complicated (i.e. with switch/case, and/or needing to encode a single byte as
an array of size 1, etc.).

Here is the diff as-is. Please feel free to use it, or to write a helper
function if you decide that's still necessary.

diff --git a/dlls/winealsa.drv/midi.c b/dlls/winealsa.drv/midi.c
index 6ef1976..359b7fc 100644
--- a/dlls/winealsa.drv/midi.c
+++ b/dlls/winealsa.drv/midi.c
@@ -932,7 +932,11 @@ static DWORD modData(WORD wDevID, DWORD dwParam)
             BYTE buf[2];
             buf[0] = evt;
             buf[1] = d1;
-            snd_seq_ev_set_sysex(&event, sizeof(buf), buf);
+            snd_midi_event_t *midi_event;
+            snd_midi_event_new(sizeof(buf), &midi_event);
+            snd_midi_event_init(midi_event);
+            snd_midi_event_encode(midi_event, buf, sizeof(buf), &event);
+            snd_midi_event_free(midi_event);
                 }
                 break;
         case 0x02:    /* Song Position Pointer. */
@@ -941,7 +945,11 @@ static DWORD modData(WORD wDevID, DWORD dwParam)
             buf[0] = evt;
             buf[1] = d1;
             buf[2] = d2;
-            snd_seq_ev_set_sysex(&event, sizeof(buf), buf);
+            snd_midi_event_t *midi_event;
+            snd_midi_event_new(sizeof(buf), &midi_event);
+            snd_midi_event_init(midi_event);
+            snd_midi_event_encode(midi_event, buf, sizeof(buf), &event);
+            snd_midi_event_free(midi_event);
                 }
             break;
         }

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.