Question: CSAP > 32 MB absread/abswrite DOES work !?!?!? (but CSAP does

Hi,
please tell me if this code that I found in SORTDIR.C of CSAP does the
handling of abswrite > 32 MB correctly. If yes, it would be an idea to
rewrite DISKIO.C in my UNDELETE tool in the same way. A32SW is a flag
that triggers "> 32 MB style" absread/abswrite there:

if (a32sw | abswrite(Disk - 'A', Dpb.ClusterSize + 1,
                              WHICHSECTOR, (char *) &Buffer[OFFSET]) {
#if 1
                a32.nsect = Dpb.ClusterSize + 1;
                a32.sector = WHICHSECTOR
;                a32.xferad = (unsigned char *) &Buffer[OFFSET];
                a32sw=1;
                if (abswrite(Disk - 'A', -1, 0, &a32) != 0) {    /* EWS */
#endif
                        fprintf(stderr, "Error writing directory.\n");
                        exit(1);
                }
#if 1
                }
#endif

a32sw is initially 0, and gets set to 1 on the first error with the old
style call.

a32 is of type:
struct absr32m
 {
  long sector;
  int  nsect;
  unsigned char far  *xferad;
 };

This looks generally promising. Please tell me if this does in fact work.
If yes, please provide a patch to MY DISKIO.C in my UNDELETE tool that
will obsolete BIGDISK.ASM :-)). Thanks a lot!

Eric


PS: This leaves the "abnormal program termination" or "endless loop" type
CSAP crashes even more unexplained! The source is in C. Maybe you can find
the bug. Ideas: Sorting of directories with 0..1 files and/or directories
and with 0 sized files could be broken in a way that does not show up in
other cases. BUT George reports that CSAP crashes almost unconditionally :-(.

Maybe it gets subdirectories wrong and only \ right? Whatever. PLEASE
be careful. If you are unlucky, CSAP will crash -after- sorting, at the
time when it rewrites the directory. This can trash your filesystem and/or
directory. Only experiment on drives that you have a backup of or if you
"know what you are doing".

Eric