Re: [Csnd] GEN16 strangeness

Date	2012-01-16 19:13
From	"Art Hunkins"
Subject	Re: [Csnd] GEN16 strangeness
	With respect to GEN16, I've discovered that it only likes extended guard point tables of 4097 and lower. It gives segmentation (or invalid pointer) errors with tables of 8193 or higher. Interestingly enough, specify a negative table length, and it's fine. (This allows for tables of arbitrary length.) For example: 8193 won't work, -8193 will. This is true for both float and double precision versions (in Windows). Below is a simple test. Could others confirm on various platforms? Art Hunkins -odac --displays -b128 -B2048 sr = 44100 ksmps = 100 nchnls = 2 instr 1 k1 linseg 0, 4, 1, 2, 1, 4, 0 k2 table k1, 3, 1 aout poscil 10000 * k2, 400, 2 outs aout, aout endin f2 0 16384 10 1 f3 0 4097 16 0 4097 .7 1; works ;f3 0 8193 16 0 8193 .7 1; doesn't work (segfault or invalid pointer) ;f3 0 -8193 16 0 8193 .7 1; works f4 0 8193 9 .25 1 0 i1 0 10 ----- Original Message ----- From: To: Sent: Monday, January 16, 2012 3:06 AM Subject: Re: [Csnd] GEN16 strangeness >I do not get a crash but i do see > PERF ERROR in instr 1: tab off end > > Using doubles and code adjusted that way > > ==John ff > > > > > Send bugs reports to the Sourceforge bug tracker > https://sourceforge.net/tracker/?group_id=81968&atid=564599 > Discussions of bugs and features can be posted here > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe > csound" > Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date	2012-01-16 20:09
From	Steven Yi
Subject	Re: [Csnd] GEN16 strangeness
	Hi Art, I tested this using the f3 that has 8193 and it is working here. I'm using the latest from git, compiled for doubles and i386/x86_64. I saw changes in git that I assume fix the issue. steven On Mon, Jan 16, 2012 at 2:13 PM, Art Hunkins wrote: > With respect to GEN16, I've discovered that it only likes extended guard > point tables of 4097 and lower. It gives segmentation (or invalid pointer) > errors with tables of 8193 or higher. > > Interestingly enough, specify a negative table length, and it's fine. (This > allows for tables of arbitrary length.) For example: 8193 won't work, -8193 > will. > > This is true for both float and double precision versions (in Windows). > > Below is a simple test. Could others confirm on various platforms? > > > Art Hunkins > > > > > > -odac --displays -b128 -B2048 > > > > > sr = 44100 > ksmps = 100 > nchnls = 2 > > instr 1 > > k1 linseg 0, 4, 1, 2, 1, 4, 0 > k2 table k1, 3, 1 > aout poscil 10000 * k2, 400, 2 > outs aout, aout > > endin > > > > > > f2 0 16384 10 1 > f3 0 4097 16 0 4097 .7 1; works > ;f3 0 8193 16 0 8193 .7 1; doesn't work (segfault or invalid pointer) > ;f3 0 -8193 16 0 8193 .7 1; works > > f4 0 8193 9 .25 1 0 > i1 0 10 > > > > > > ----- Original Message ----- From: > To: > Sent: Monday, January 16, 2012 3:06 AM > > Subject: Re: [Csnd] GEN16 strangeness > > >> I do not get a crash but i do see >> PERF ERROR in instr 1: tab off end >> >> Using doubles and code adjusted that way >> >> ==John ff >> >> >> >> >> Send bugs reports to the Sourceforge bug tracker >> https://sourceforge.net/tracker/?group_id=81968&atid=564599 >> Discussions of bugs and features can be posted here >> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe >> csound" >> > > > > Send bugs reports to the Sourceforge bug tracker > https://sourceforge.net/tracker/?group_id=81968&atid=564599 > Discussions of bugs and features can be posted here > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe > csound" > Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date	2012-01-16 20:34
From	Tito Latini
Subject	Re: [Csnd] GEN16 strangeness
Attachments	None

Date	2012-01-16 21:05
From	Steven Yi
Subject	Re: [Csnd] GEN16 strangeness
	Hi Art, Sorry that was confusing, I tested on OSX 10.7. Thanks! steven On Mon, Jan 16, 2012 at 3:09 PM, Steven Yi wrote: > Hi Art, > > I tested this using the f3 that has 8193 and it is working here. I'm > using the latest from git, compiled for doubles and i386/x86_64. I > saw changes in git that I assume fix the issue. > > steven > > On Mon, Jan 16, 2012 at 2:13 PM, Art Hunkins wrote: >> With respect to GEN16, I've discovered that it only likes extended guard >> point tables of 4097 and lower. It gives segmentation (or invalid pointer) >> errors with tables of 8193 or higher. >> >> Interestingly enough, specify a negative table length, and it's fine. (This >> allows for tables of arbitrary length.) For example: 8193 won't work, -8193 >> will. >> >> This is true for both float and double precision versions (in Windows). >> >> Below is a simple test. Could others confirm on various platforms? >> >> >> Art Hunkins >> >> >> >> >> >> -odac --displays -b128 -B2048 >> >> >> >> >> sr = 44100 >> ksmps = 100 >> nchnls = 2 >> >> instr 1 >> >> k1 linseg 0, 4, 1, 2, 1, 4, 0 >> k2 table k1, 3, 1 >> aout poscil 10000 * k2, 400, 2 >> outs aout, aout >> >> endin >> >> >> >> >> >> f2 0 16384 10 1 >> f3 0 4097 16 0 4097 .7 1; works >> ;f3 0 8193 16 0 8193 .7 1; doesn't work (segfault or invalid pointer) >> ;f3 0 -8193 16 0 8193 .7 1; works >> >> f4 0 8193 9 .25 1 0 >> i1 0 10 >> >> >> >> >> >> ----- Original Message ----- From: >> To: >> Sent: Monday, January 16, 2012 3:06 AM >> >> Subject: Re: [Csnd] GEN16 strangeness >> >> >>> I do not get a crash but i do see >>> PERF ERROR in instr 1: tab off end >>> >>> Using doubles and code adjusted that way >>> >>> ==John ff >>> >>> >>> >>> >>> Send bugs reports to the Sourceforge bug tracker >>> https://sourceforge.net/tracker/?group_id=81968&atid=564599 >>> Discussions of bugs and features can be posted here >>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe >>> csound" >>> >> >> >> >> Send bugs reports to the Sourceforge bug tracker >> https://sourceforge.net/tracker/?group_id=81968&atid=564599 >> Discussions of bugs and features can be posted here >> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe >> csound" >> Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"