> Please try http://www.fdos.org/kernel/testing/devload-3.23.zip
> and let me know if you experience any issues with it.
>
> I have done limited testing with it using shsucdx and tdsk and it
> seems to fix the issue.  Note that if you install shsucdx then tdsk
> then uninstall shsucdx thus creating a hole in the used drives,
> installing a second tdsk with devload will use the drive letter of the
> hole.  I have not tested with subst drives, but if there are issues
> then the flag check may need adjusted.

I would change the code at around line 545, where it uses the numBlockDev
value to index into the CDS. It isn't consistent to expect all block
devices to be consecutive at the beginning of the CDS. As you explained,
that isn't always the case. The way numBlockDev is used here though the
loop might not ever check some of the gaps.

(You could implement a scheme where it scans the top part of the CDS first
and then tries the lower part, but I do not see much benefit in that.
However, user-settable drive assignments would be interesting, ie,
providing command line options to specify which drives are tried. I think
SHSUCDX implements some such options. There's no technical reason DEVLOAD
couldn't.)

Regardless, there are also two bugs in the loop:

* It assumes that at least one entry is free. It will scan beyond the CDS
if that is not the case.

* It assumes that numBlockDev is lower than numCDSEntries.

Both can be fixed by inserting a check immediately after the CDSinuse
label to insure that the next entry is not yet beyond the CDS. (The check
could be inserted in front of CDSfound, but that wouldn't fix the second
bug.)

Regards,
Christian

------------------------------------------------------------------------------
BlackBerry® DevCon Americas, Oct. 18-20, San Francisco, CA
The must-attend event for mobile developers. Connect with experts. 
Get tools for creating Super Apps. See the latest technologies.
Sessions, hands-on labs, demos & much more. Register early & save!
http://p.sf.net/sfu/rim-blackberry-1
_______________________________________________
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel

Reply via email to