This could explain why I can't get LSL3 save games to work now. Have you noticed
this Matt?
I haven't yet changed the logic behind sci0_soundserver() (unless I made a
mistake somewhere in my last commit), and quite frankly have pretty much decided
not to. :)
Christoph - because some extra variables will need to be saved in games when the
event sound server comes along, is it OK to put these in and break everything or
do you have some good suggestions?
Just finished exams today so will be able to work on this again soon.
Cheers,
Alex.
> Trying to debug the CB1 problem on win32, and am noticing the following
> (new) purify warnings:
>
> [E] ABR: Array bounds read in sci0_soundserver {33 occurrences}
> Reading 1 byte from 0x02c56110 (1 byte at 0x02c56110 illegal)
> Address 0x02c56110 is 1 byte past the end of a 64 byte block at
> 0x02c560d0
> Address 0x02c56110 points to a malloc'd block in heap 0x01e10000
> Thread ID: 0x56c
> Error location
> sci0_soundserver [soundserver.c:224]
> ** at this point */
>
> /* Handle length escape sequence (see SCI
> sound specs) */
> => while ((tempticks = song->data[(song->pos)++])
> == SCI_MIDI_TIME_EXPANSION_PREFIX)
> ticks += SCI_MIDI_TIME_EXPANSION_LENGTH;
> ticks += tempticks;
> }
>
>
> [W] UMR: Uninitialized memory read in gfxwop_dyn_view_compare_to {1
> occurrence}
> Reading 4 bytes from 0x02d2f8dc (4 bytes at 0x02d2f8dc
> uninitialized)
> Address 0x02d2f8dc is 156 bytes into a 164 byte block at
> 0x02d2f840
> Address 0x02d2f8dc points to a malloc'd block in heap 0x01e10000
> Thread ID: 0x368
> Error location
> gfxwop_dyn_view_compare_to [widgets.c:983]
> if (retval)
> return retval;
>
> => return -(wview->sequence - oview->sequence);
> }
>
>
>
>
>
> --
> http://www.clock.org/~matt
>
>
