Hi,
i had a look at _cdio_get_track_msf in lib/driver/MSWindows, because i
assumed that this would be used with MinGW.
The driver seems only partly prepared for start track number > 1.
But i cannot derive the particular mistake of cd-info from what i see in
the code.
So which driver is in effect with that failed test ?
-------------------------------------------------------------------------
What's wrong with lib/driver/MSWindows:
The implementation of .get_track_msf does not subtract the track number
offset, but rather subtracts 1 before using its parameter i_tracks as
index to p_env->tocent[].
...
The "leadout" track is specified either by
using i_tracks LEADOUT_TRACK or the total tracks+1.
...
*/
static bool
_cdio_get_track_msf(void *p_user_data, track_t i_tracks, msf_t *p_msf)
{
...
if (i_tracks == CDIO_CDROM_LEADOUT_TRACK) i_tracks = p_env->gen.i_tracks+1;
if (i_tracks > p_env->gen.i_tracks+1 || i_tracks == 0) {
return false;
} else {
cdio_lsn_to_msf(p_env->tocent[i_tracks-1].start_lsn, p_msf);
return true;
}
The two implementations of reading the Table-Of-Content subtract the
track number offset when filling p_env->tocent.
lib/driver/MSWindows/aspi32.c : read_toc_aspi
p_env->gen.i_tracks = tocheader[3] - tocheader[2] + 1;
...
for( i = 0 ; i <= p_env->gen.i_tracks ; i++, j++ ) {
...
p_env->tocent[ i ].start_lsn = ...
lib/driver/MSWindows/win32_ioctl.c : read_toc_win32ioctl
p_env->gen.i_tracks = cdrom_toc.LastTrack - cdrom_toc.FirstTrack + 1;
...
for( i = 0 ; i <= p_env->gen.i_tracks ; i++, j++ ) {
p_env->tocent[ i ].start_lsn =
So to deliver the correct entries, _cdio_get_track_msf() has to be called
with i_tracks = tocent index + 1. I.e. first valid i_tracks is 1, not
the nominal first track number.
But src/cd-info.c calls cdio_get_track_msf and thus win32.c
_cdio_get_track_msf with indice from a loop starting at the nominal
first track number:
for (i = i_first_track; i <= CDIO_CDROM_LEADOUT_TRACK; i++) {
...
if (!cdio_get_track_msf(p_cdio, i, &msf)) {
...
...
/* skip to leadout */
if (i == i_first_track + i_tracks - 1) {
i = CDIO_CDROM_LEADOUT_TRACK-1;
}
}
So it seems not much of a riddle why the leadout LSN and MSF are wrong.
But why then is the cd-info line for track 5 correct ?
Why does win32.c:_cdio_get_track_msf not bail out by this test
when i_tracks is 4 or 5 ? (p_env->gen.i_tracks is supposed to be 2.)
if (i_tracks > p_env->gen.i_tracks+1 || i_tracks == 0) {
return false;
-------------------------------------------------------------------------
Have a nice day :)
Thomas
