At 10:04 PM 9/7/2004 -0500, I wrote:
I have an alternate plan. If I still have and can find one of the many MS-DOS disks I owned for now defunct machines, or otherwise acquire a Bochs-style disk image of a true MS-DOS 6.x kernel, I could check the ZRDX failure path in QEMU trace log and see if it fails the same way. If so, we know 4ch behaves differently. If not, the problem lies elsewhere. Could be a ZRDX bug either way, though.
Alright, ZRDX version 0.49 which comes bound to Kaspersky anti-virus is buggy, because there is also a failure exit under true MS-DOS, although MS-DOS doesn't lock like FreeDOS. FreeDOS failed at a different point at this test, meaning there is general instability in that environment.
Binding up to ZRDX version 0.50 allows Kaspersky to run under my machine and successfully exit to FreeDOS command prompt, as does DOS/4GW and CauseWay. I will update Bugzilla to reflect these findings.
As a side-note: true MS-DOS 6.22 turns off both the cache disable and no write-through bits of the CPU (>=80486) from their reset state of on, either in its kernel or HIMEM.SYS. FreeDOS might be able to do that and get better overall system performance since CPU caching is disabled by default. I don't know if there are any serious repercussions to doing this and how much of a measured performance improvement there might be. If there are benchmark people around who can get a good compare, a test would be simple since CR0 need only be AND'ed with NOT 0x60000000 to turn them off.
------------------------------------------------------- This SF.Net email is sponsored by: thawte's Crypto Challenge Vl Crack the code and win a Sony DCRHC40 MiniDV Digital Handycam Camcorder. More prizes in the weekly Lunch Hour Challenge. Sign up NOW http://ad.doubleclick.net/clk;10740251;10262165;m _______________________________________________ Freedos-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-devel
