Re: [hatari-devel] TT emulation crashes when there is no ACSI drive

[Uwe Seimet <Uwe.Seimet@xxxxxxxxx>]

The crash happens even before any code is executed. This is the
stripped-down main method:

int
main(WORD argc, const char *argv[])
{
        printf("\ntest0\n");

        Cconin();

        return 0;
}

The printf is not reached, the bus error happens before. The problem
might even be reproducable on Windows or a Mac, because there is no
NatFeats call at all. In my AUTO folder the file is called 1.PRG instead
of NF_SCSI.PRG, in order for it to be started first.
When there is an ACSI device the test output is printed and I have to
press a key, just as you would expect.

See the attachment with the binary of this program with the stripped
main method pasted above.

> I'll try to identify the NF_SCSI call that triggers this.
> 
> > Le 11/08/2019 à 10:33, Christian Zietz a écrit :
> > > Nicolas Pomarède schrieb:
> > > 
> > >> looking at the code, unlike what the commit's name imply, this is mostly
> > >> about natfeat (which was merged with new MFP code).
> > > 
> > > I think the Tuxfamily Git web interface is misleading here because it
> > > doesn't properly interpret the merge (without rebasing first) that you did.
> > > 
> > > Running "git diff 65dda3cf 50bf142e" shows you the true difference
> > > between the two versions that Uwe tested. A diff is attached for easier
> > > reference. It's indeed about the TT MFP.
> > 
> > Hi
> > 
> > I don't see anything obvious or really ACSI related here, so hard to 
> > tell, especially if no one is able to reproduce the problem to get a 
> > starting base :(
> > 
> > (still, I think there's a bug in nafteat code as pointed in previous 
> > mail, even if not related to this)
> > 
> > Nicolas
> > 
> > 
> 
>

Attachment: 1.PRG
Description: Binary data