Re: [hatari-devel] Automounting HDD images

[Thomas Huth <th.huth@xxxxxx>]

Am Tue, 21 Jan 2014 22:51:39 +0100
schrieb Miro Kropáček <miro.kropacek@xxxxxxxxx>:

> > How did you determine this?
> >
> > If Hatari would cache the whole image content, that could require
> > more memory than the host machine has...
> >
> > At least hdc.c seems to do normal fseek & fread/fwrite when
> > accessing the drive content.
> >
> I'm not saying I know why but this is what happens. Even if I 'Eject'
> the image and browse to the same filename, I'm still getting the old
> image content (tested with bootsector messages). Maybe it's a Linux
> configuration issue, I don't really know but I have no clue how to
> fix this.

Could be a stdio buffering issue... you could try to add a
"setvbuf(fhndl, NULL,  _IONBF, 0);" right after the corresponding
fopen() call to disable stdio buffering to see whether it makes a
difference...?

 Thomas