From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Fjellstrom Subject: Re: rawmidi problem (was Re: ALSA homepage redesign) Date: Tue, 23 Apr 2002 17:43:43 -0600 Sender: alsa-devel-admin@lists.sourceforge.net Message-ID: <0GV100DRBOG2ST@l-daemon> References: Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 7BIT Return-path: In-Reply-To: Errors-To: alsa-devel-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Unsubscribe: , List-Archive: To: Takashi Iwai Cc: alsa-devel@lists.sourceforge.net List-Id: alsa-devel@alsa-project.org On April 23, 2002 05:44 am, you wrote: > At Tue, 23 Apr 2002 12:14:42 +0200, > > I wrote: > > At Mon, 22 Apr 2002 13:50:27 -0600, > > > > Thomas Fjellstrom wrote: > > > I don't hear any sound when playing midi using the rawmidi interface > > > (using my SBLive! Value), I tried, some code written by me, and the > > > rawmidi.c source demo that is included with the driver package. (A > > > sound font is loaded, and midis play fine with sound, using the pmidi > > > program.) > > > > hmm it looks like a bug was introduced in virmidi or in a combination > > of virmidi and rawmidi. i think the rawmidi itself works fine. > > this problem appears only on emu10k1's virmidi. > > > > i'll take a further look at this problem. > > now fixed on cvs. actually it was a bug on virmidi. > > thanks for your notice. > No problem, except now I've noticed a few things :) One: SND_RAWMIDI_SYNC isn't a valid flag for snd_rawmidi_open. ( rawmidi_hw.c:221: snd_rawmidi_hw_open: Assertion `!(mode & ~(0x0001|0x0002))' failed.) Two: using SND_RAWMIDI_APPEND on snd_rawmidi_open causes an assert then a segfault.. A '0' for flags is fine... (see Three) rawmidi.c:963: snd_rawmidi_write: Assertion `rawmidi' failed. (this error is printed thousands of times... before a segfault) Three: after a while, about a minute or so of playing a midi, the midi starts getting slower and slower.. and CPU usage goes to 0... I'm sure Its not the midi player, as I've tried it's 'DIGIMID' driver (which uses a converted soundfont to play through the digital pcm interface) and it doens't have this problem. (now I've got to reboot... because I can't think of another way to turn off the midi notes that were left on... :) -- Thomas Fjellstrom tfjellstrom@shaw.ca http://strangesoft.net