Hi all,

I've been working on my fluid wrapper tonight to get rid of underscores
and I found that csound5 was almost altogether busted for me, not
running in deferred or realtime.  I ran with --verbose flag and got this
last statement:

argoff array:
        0       1       24      1       2       1       25      1      
2      26       1       3       1       27      1       4       1      
37      1      38       1       5       1       39      1       5      
1       32      1      33       2       1       32      5       32     
33      33      5       5      34       2       1       32      5      
32      34      34      5       5      35       2       1       32     
5       32      35      35      5       5      36       2       1      
32      5       32      36      36      5       5      -9       1      
-4      1       -10     1       -5      1       -11     2      -9      
6       1       -12     1       -9      1       -13     3       -11   
-12      -9      1       -9      1       -13     1       -14     2      
-10    -15      1       -10     1       -16     3       -14     -15    
-10     1      -10      1       -16     1       -17     3       -9     
-3      -10     1      -18      1       -6      1       -19     1      
-18     1       -20     1      -7       1       -21     1       -8     
1       -22     1       5       1      -23      2       -9      -10    
1       -24     3       -23     -9      -10    -25      1       -24    
1       -26     2       -21     5       2       -26    -32768   1      
-27     2       -19     4       1       -28     2       -3     -29     
2       -19     4       1       -30     7       5       9       -29   
-28      -27     8       5       1       -63     7       -30     -17    
-25    -64      5       -63     11      10      5       5       1      
-65     5      -63      13      12      5       5       1       -66    
5       -63     14     10       5       5       1       -67     2      
15      -64     1       -68    -67      -65     1       -69     2      
15      -66     1       -70     2      -68      -69     1       -71    
2       16      -63     1       -72     2      -70      -71     1      
-73     4       -72     -63     17      5       1      -31      2      
-21     3       2       -31     -32767  1       -32     2      -19     
4       1       -33     2       -3      7       1       -34     2     
-19      4       1       -30     7       5       9       -34     -33    
-32    -35      2       3       -7      1       -36     2       -35    
-3      1      -37      2       -36     18      1       -38     2      
-7      -3      1      -39      7       -5      -38     -5      18     
-6      -37     -6      1      -63      7       -30     -39     -5     
5       3       5       5       1      -64      5       -63     11     
10      5       5       1       -65     5      -63      13      12     
5       5       1       -66     5       -63     14     10       5      
5       1       -74     2       15      -64     1       -75    -74     
-65     1       -76     2       15      -66     1       -77     2     
-75      -76     1       -78     2       16      -63     1       -79    
2      -77      -78     1       -73     4       -79     -63     17     
5       1      -40      2       -21     4       2       -40     -32766 
1       -41     2      -19      4       1       -42     2       -3     
19      1       -43     2      -19      4       1       -30     7      
5       9       -43     -42     -41    20       5       1       -73    
7       -30     -5      -5      5       21     -44      2       -20    
22      1       -45     2       -44     23      1      -20      1      
-45     1       -46     1       4       1       -47     2      -46     
4       1       -48     1       -20     1       -49     1       -20   
-50      2       -48     -49     1       -51     2       -47     -50    
1      -52      1       -51     1       -53     1       4       1      
-54     2      -53      4       1       -55     1       -20     1      
-56     1       -20    -57      2       -55     -56     1       -58    
2       -54     -57     1      -59      1       -58     1       -80    
2       -73     -52     1       -81    -80      1       -82     2      
-73     -59     1       -83     1       -82    -84      2       37     
-81     1       37      1       -84     1       -85    38       -83    
1       38      1       -85     4       32      33      -4     -5      
4       32      34      -4      -5      4       32      35      -4    
-5       4       32      36      -4      -5      2       -4      -5     
1      32       1       -6      2       -4      28      1       -4     
1       -6     -7       2       -5      28      1       -5      1      
-7      2       -4     -5
Csound tidy up: Segmentation fault

I'd like to run in gdb but wasn't sure what would be the best way to go
about it (I don't use gdb very regularly).  When I'm running gdb and get
a segfault like this, what command should I use to figure out where
things died?

Also, a separate question, what do using 0xffff, 0xfffe, 0xfffd, and
0xfffc mean for dsblksize in an OENTRY?  I came across this in entry2.c
while trying to remember parameters for OENTRY and I realized I never
understood exactly what this meant.  

Thanks,
steven

>>>>> "stevenyi" == stevenyi   writes:

 stevenyi> Also, a separate question, what do using 0xffff, 0xfffe, 0xfffd, and
 stevenyi> 0xfffc mean for dsblksize in an OENTRY?  I came across this in entry2.c
 stevenyi> while trying to remember parameters for OENTRY and I realized I never
 stevenyi> understood exactly what this meant.  

Answering the easy question first; in OENTRY the size field is used as
a flag in saying what types of polymorphism are allowed.

0xffff means that the first argument type determines which opcode is
actually being used, XXX -> XXX_a or XXX_i or XXX_k

0xfffe means that the first two argument types determine the actual
opcode.  So XXX -> XXX_aa, XXX_ak, XXX_ka, and XXX_kk.  Think oscil

0xfffd looks at one argument but does not allow i type (eg peak)

0xfffc is rather like 0xfffe but deals with divz -- so allows a, k and
i arguments

This is "clear" from the header in entry1.c and the code in rdorch.c

==John ff

This is a rewrite of Barry's original that changed XXX to aaXXX etc
and caused inadvertent name clashes.  The divz and peak variants are
entirely mine.  Rewrite took place one Sunday in Somerset,Ma.  Blame
the view!










==John ffitch

>>>>> "stevenyi" == stevenyi   writes:

 stevenyi> Hi all,
 stevenyi> I've been working on my fluid wrapper tonight to get rid of underscores
 stevenyi> and I found that csound5 was almost altogether busted for me, not
 stevenyi> running in deferred or realtime.  I ran with --verbose flag and got this
 stevenyi> last statement:

the --verbose arg ensures that this array is printed.  It is not very
useful unless you are looking at args

 stevenyi> argoff array:
 stevenyi>         0       1       24      1       2       1       25      1      
 stevenyi> 2      26       1       3       1       27      1       4       1      
etc etc 

 stevenyi> Csound tidy up: Segmentation fault

On the other hand this says you have a segmentation error!

Without knowing the orchestra/platform it is hard to say much other
than run under gdb (Linux and OSX anyway) and go "where" when it
breaks.  That is my usual procedure in these cases.  The trace before
might have given some clues.

==John ffitch

How can I produce unix style line breaks on windows? I've looked at 
Dev-C++, emacs and vim and didn't find and 'evident' way of doing it.
Thanks
Andrés

> How can I produce unix style line breaks on windows?
If you use emacs the commands "C-X Ret f undecided-unix" do it with ease
==John ffitch

Hi John,

Thanks for the explanation!  That clears it up for me just nicely.  I'm 
going to add that to the csound.tex file in just a moment to keep that 
information handy.

steven



jpff@codemist.co.uk wrote:
>>>>>>"stevenyi" == stevenyi   writes:
> 
> 
>  stevenyi> Also, a separate question, what do using 0xffff, 0xfffe, 0xfffd, and
>  stevenyi> 0xfffc mean for dsblksize in an OENTRY?  I came across this in entry2.c
>  stevenyi> while trying to remember parameters for OENTRY and I realized I never
>  stevenyi> understood exactly what this meant.  
> 
> Answering the easy question first; in OENTRY the size field is used as
> a flag in saying what types of polymorphism are allowed.
> 
> 0xffff means that the first argument type determines which opcode is
> actually being used, XXX -> XXX_a or XXX_i or XXX_k
> 
> 0xfffe means that the first two argument types determine the actual
> opcode.  So XXX -> XXX_aa, XXX_ak, XXX_ka, and XXX_kk.  Think oscil
> 
> 0xfffd looks at one argument but does not allow i type (eg peak)
> 
> 0xfffc is rather like 0xfffe but deals with divz -- so allows a, k and
> i arguments
> 
> This is "clear" from the header in entry1.c and the code in rdorch.c
> 
> ==John ff
> 
> This is a rewrite of Barry's original that changed XXX to aaXXX etc
> and caused inadvertent name clashes.  The divz and peak variants are
> entirely mine.  Rewrite took place one Sunday in Somerset,Ma.  Blame
> the view!
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> ==John ffitch
> 
> 
> 

Hi John,

Thanks for the advice on 'where'!  I'll give it a try and hope debugging 
won't end up an all night affair.  =)

steven


jpff@codemist.co.uk wrote:


> Without knowing the orchestra/platform it is hard to say much other
> than run under gdb (Linux and OSX anyway) and go "where" when it
> breaks.  That is my usual procedure in these cases.  The trace before
> might have given some clues.
> 
> ==John ffitch
> 
> 
>