On Wed, 24 Mar 2004, tom ehlert wrote: > So I recommend setting the default to > stacks=8,256 > > which the user may overwrite as he likes.
OK. Too bad but not much we can do about that. > > >Now the code (HMATEXT) overhead is ~300 bytes, plus we have those two low > >fnodes that take 120 bytes together. But with say 30 files 1800 bytes > >could be moved to UMBs. > > for some reasons, I don't use EMM386 (some network cards cause trouble) > > so I don't have upper memory, so there is no gain. > > However the internal fnodes could go to HMA, then there would be a > real gain, even for me ;) The far fnodes could go to the HMA; the two low fnodes not (without code size increase from cs: prefixes and tricky code). However. Much of the information in the far fnodes is also in the SFTs. Instead of fmemcpy(near fnode, far fnode) a copy from SFT fields to fnode fields can be done now without affecting the main fatfs code. This way the far fnodes can be greatly reduced in size or even completely eliminated, benefiting everyone in the end. Bart ------------------------------------------------------- This SF.Net email is sponsored by: IBM Linux Tutorials Free Linux tutorial presented by Daniel Robbins, President and CEO of GenToo technologies. Learn everything from fundamentals to system administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click _______________________________________________ Freedos-kernel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-kernel
