Re: [hylafax-users] Send program terminated abnormally with exit status 0xa

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Xavier,

I'm glad that the code change helped.  I'm not sure, however, that it 
gives me any conclusive information as far as *what* is wrong.

In the problem scenario...

   for (u_short fnum = 0; fnum < frameNumber; fnum++) {
       ... stuff ...
   }

frameNumber here *is supposed to be* 256.  So by changing that to...


   for (u_short fnum = 0; fnum < frameNumber && fnum < 256; fnum++) {
       ... stuff ...
   }

is just being redundant (fnum < 256 && fnum < 256).  It, however, would 
seem to indicate that frameNumber *is not* 256.  Yet, there also is the 
problem that here your system also thinks that (u_short) 255 + 1 = 
2336.  So, I'm not sure if we're looking at a problem with fnum, 
frameNumber, or both.

If your compiler does limit u_short to 8 bits (and that's not clearly 
evident because fnum was shown to contain a value larger than 255), then 
(your compiler, frankly, is broken) we'd have to scour the source code 
to make sure that no other shorts can get that big.  There may not be 
any other cases of it other those that deal with frame numbering.  But, 
I'm more inclined to let you pursue this with Apple, as it's very 
evident (255 + 1 = 2336) that something's wrong with the compiler.

Lee.


Xavier Blanco wrote:

> Hi Lee,
>
> I have done your suggested changes in source code, recompille and reinstall
> and now I am able to send faxes without crashing faxsend :-)
>
> But reading your words I understand that I can find more crashing code due
> my compiler is not assigning the correct bit length to short type. Perhaps
> some option at compile time is needed ?
>
> Thank you
>
> regards
>
> xavier
>
>
>  
>
> > De: Lee Howard <faxguy@xxxxxxxxxxxxxxxx>
> > Fecha: Tue, 15 Mar 2005 15:57:44 -0800
> > Para: xavier@xxxxxxxxxxxxx
> > CC: hylafax-users@xxxxxxxxxxx
> > Asunto: Re: [hylafax-users] Send program terminated abnormally with exit
> > status 0xa
> >
> > Xavier Blanco wrote:
> >
> >    
> >
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 251
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 252
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 253
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 254
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 255
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 2336
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 2337
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 2338
> > > Mar 15 17:59:16.86: [ 2014]: SEND send frame number 2340
> > >
> > >
> > >      
> > This is a problem.  There can only be frames up to number 255.
> >
> > The code responsible here is faxd/Class1Send.c++ at approximately line
> > 955.  The number mentioned there in the logs ("fnum" in the code)is
> > defined as type u_short.  It does this:
> >
> > for (u_short fnum = 0; fnum < frameNumber; fnum++) {
> > ... stuff ...
> > }
> >
> > Somehow it thinks that 255+1 = 2336
> >
> > As you're using a platform (and probably a compiler) that is largely
> > untested (OS X 10.3.8 is rather new), I'm guessing that the compiler
> > doesn't like u_short.
> >
> > According to the information I have, we're *supposed* to have 16 bits
> > with a short type.  As 0-255 only consume 8 bits, I can only suppose
> > that your compiler is only allocating 8 bytes to shorts.  If this is
> > correct then this could be a problem throughout the code, as HylaFAX
> > uses shorts quite frequently.
> >
> > You could do a couple of tests, perhaps.  Change the code to read...
> >
> > for (u_short fnum = 0; fnum < frameNumber && fnum < 256; fnum++) {
> > ... stuff ...
> > }
> >
> > Lee.
> >
> >    
>
>
> ____________________ HylaFAX(tm) Users Mailing List _______________________
>  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
> On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
>  *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*
>
>  


____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*