[Cs-dev] new bugs in turnoff2 and diskin

[Cs-dev] new bugs in turnoff2 and diskin

Date	2007-09-30 05:31
From	Jonathan Murphy
Subject	[Cs-dev] new bugs in turnoff2 and diskin
	I hadn't updated my version until yesterday, so I'm sorry that this is last minute. It seems that the changes to goto_ops.c and diskin.c have introduced problems. I think that the change to goto_ops.c has broken turnoff2 for one or several modes. Reverting to the previous version fixes the issues, so the new code is definitely broken. I haven't explicitly checked diskin yet, but I'm pretty sure that the new buffersize parameter is the culprit. I'll rebuild later today and get debugging info. Jonathan. ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net

Date	2007-09-30 10:42
From	Jonathan Murphy
Subject	Re: [Cs-dev] new bugs in turnoff2 and diskin
	I've worked out what's going on. diskin is fine, so sorry about the noise there. The problem is just with turnoff2, in the case of mode 4, when the instrument has already been deactivated. Instead of returning a warning, csound hangs. Here is a patch: --- goto_ops.c.1.9 2007-09-30 18:24:32.000000000 +0900 +++ goto_ops.c 2007-09-30 18:29:21.000000000 +0900 @@ -187,6 +187,9 @@ } ip = &(csound->actanchor); ip2 = NULL; + if ((mode & 4) && !ip->p1){ + return csoundPerfError(csound, Str("turnoff2: invalid instrument number")); + } while ((ip = ip->nxtact) != NULL && (int) ip->insno != insno); if (ip == NULL) return OK; Jonathan. Jonathan Murphy writes: > I hadn't updated my version until yesterday, so I'm sorry that this is > last minute. It seems that the changes to goto_ops.c and diskin.c have > introduced problems. I think that the change to goto_ops.c has broken > turnoff2 for one or several modes. Reverting to the previous version > fixes the issues, so the new code is definitely broken. I haven't > explicitly checked diskin yet, but I'm pretty sure that the new > buffersize parameter is the culprit. I'll rebuild later today and get > debugging info. > > Jonathan. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Csound-devel mailing list > Csound-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net

Date	2007-09-30 12:50
From	Jonathan Murphy
Subject	Re: [Cs-dev] new bugs in turnoff2 and diskin
	Sorry, that patch won't work. Jonathan. Jonathan Murphy writes: > I've worked out what's going on. diskin is fine, so sorry about the > noise there. The problem is just with turnoff2, in the case of mode 4, > when the instrument has already been deactivated. Instead of returning > a warning, csound hangs. Here is a patch: > > > --- goto_ops.c.1.9 2007-09-30 18:24:32.000000000 +0900 > +++ goto_ops.c 2007-09-30 18:29:21.000000000 +0900 > @@ -187,6 +187,9 @@ > } > ip = &(csound->actanchor); > ip2 = NULL; > + if ((mode & 4) && !ip->p1){ > + return csoundPerfError(csound, Str("turnoff2: invalid instrument number")); > + } > while ((ip = ip->nxtact) != NULL && (int) ip->insno != insno); > if (ip == NULL) > return OK; > > Jonathan. > > Jonathan Murphy writes: > >> I hadn't updated my version until yesterday, so I'm sorry that this is >> last minute. It seems that the changes to goto_ops.c and diskin.c have >> introduced problems. I think that the change to goto_ops.c has broken >> turnoff2 for one or several modes. Reverting to the previous version >> fixes the issues, so the new code is definitely broken. I haven't >> explicitly checked diskin yet, but I'm pretty sure that the new >> buffersize parameter is the culprit. I'll rebuild later today and get >> debugging info. >> >> Jonathan. >> >> >> ------------------------------------------------------------------------- >> This SF.net email is sponsored by: Microsoft >> Defy all challenges. Microsoft(R) Visual Studio 2005. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> Csound-devel mailing list >> Csound-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/csound-devel > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Csound-devel mailing list > Csound-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net

Date	2007-09-30 15:23
From	jpff
Subject	Re: [Cs-dev] new bugs in turnoff2 and diskin
	True! ==John ffitch ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net

Date	2007-09-30 16:57
From	Jonathan Murphy
Subject	Re: [Cs-dev] new bugs in turnoff2 and diskin
	However, this csd is broken by the changes to turnoff2. It works if I revert to the previous version. Any suggestions? Jonathan. -+rtaudio=alsa -odac -+rtmidi=alsa -Mhw:1,0 -d -m0 sr = 48000 ksmps = 16 nchnls = 2 0dbfs = 1 massign 0, 0 pgmassign 0, 0 gisin ftgen 1, 0, 32768, 10, 1 giji_12 ftgen 2, 0, 32, -2, 12, 2, 256, 60, 1, \ 16/15, 9/8, 6/5, 5/4, 4/3, 7/5, 3/2, 8/5, 5/3, 9/5, 15/8, 2 gkon init 0 gkmax init 8 instr 1 kst, kch, kd1, kd2 midiin if (kst == 144) then kdec = kd1/1000 kin = 2.0 + kdec if (kd2 == 0) then turnoff2 kin, 4, 1 gkon = gkon - 1 elseif (kd2 > 0) then event "i", kin, 0, -1, kd1, kd2 gkon = gkon + 1 if (gkon > gkmax) then turnoff2 2, 1, 1 gkon = gkon - 1 endif endif endif endin instr 2 ikey = p4 ivel = p5 * 0.0078125 icps cpstuni ikey, giji_12 iatt = 0.03 idec = 10 isus = 0 irel = 1 xtratim irel krel release if (krel == 1) kgoto rel aenv1 linseg 0, iatt, 1, idec, isus, -1, isus aenv = aenv1 kgoto done rel: aenv2 linseg 1, irel, 0, -1, 0 aenv = aenv1 * aenv2 done: aosc oscil ivel, icps, gisin aosc = aosc * aenv * 0.2 outs aosc, aosc endin i1 0 10000 e jpff writes: > True! > ==John ffitch > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Csound-devel mailing list > Csound-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net