SweetLow

Release: https://github.com/LordOfMice/Tools/blob/master/ramdrv4m.zip >Unresolved problems at this time: >1. Due how the IOS works, there is two consecutive >blue screens with warning on exit to DOS when a swap file is located >on RAM Drive. This behavior is non-specific to RAMDRV4M. >2. RLOEW disks do not work after exiting to DOS Fixed too early destruction of controller object on Windows exit.

Release: https://github.com/LordOfMice/Tools/blob/master/cregfix.zip Added .SYS and DPMI versions

JFYI. much modern version of this pack is available now: http://sweetlow.orgfree.com/download/usb20_win9x.zip

Version for testing: http://sweetlow.orgfree.com/download/ramdrv4m.zip RAMDRV4M->README: >Unresolved problems at this time: >1. Due how the IOS works, there is two consecutive >blue screens with warning on exit to DOS when a swap file is located >on RAM Drive. This behavior is non-specific to RAMDRV4M. >2. RLOEW disks do not work after exiting to DOS. I managed to write a workaround for these problems (under Windows 95 & 98, not ME) and it turned out that both problems are actually the same problem - unplanned (and too early) deletion of the drive by the OS. However, the source of such impolite behavior is unknown still.

http://sweetlow.orgfree.com/download/cregfix.zip For test: Added WP_OFF.EXE and as it is PM application it clears CR0.WP (Write Protect) bit only. Works with Ring 0 DPMI hosts only - as included CWSDPMI R5 and R7 versions.

I assume just because it is relatively easy way to deal with PM environment.

It is good logic, but this problem now is minor for systems with this bug. Ok, but i think it is not so simple for DPMI executables at all.

And why did you ask solution without having problem?

N.B. And probably you can just do such version youself if you need it: COPY /B CREGFIX.EXE+CREGFIX.SYS CREGFIX.EXE, isn't it?

It is only half of test. Do Windows load successfully after that?

Two solution exist - or make driver (.SYS version) and run it before EMM386 or make DPMI version of CREGFIX. P.S. http://sweetlow.orgfree.com/download/cregfix.zip .SYS version for test

Or limit XMS memory to 512M or use RLoew's PATCHMEM. And "MaxPhysPage setting in system.ini" is bad idea. P.S. It is pinned topic...

You should not remove but RELOCATE this item from old to new addresses: BF800000 - BFFFFFFF

It is not enough good tool in THIS area: Base Addresses: (0) D0000008 = mem base=D0000000 len=128M prefetchable (1) 0000C001 = I/O base=0000C000 len=256 (2) E5000000 = mem base=E5000000 len=64K CardBus: 00000000 ExpansionROM: 00000000 (128K,disabled)

Looks like BIOS reserved 512M for this card as 256M card usually takes 768M of address space on PCI-E systems. But show any (good) DOS PCI dumper info for this card to know it exactly.

JFYI: schwups, OS:ME

Check and try to find something not included in my collection Probably no. But you can ask him here or on old-games or on ru-board, of course. Unofficial stuff has one unsolvable problem - it depends too hard on taste of maker of the pack (and third party software he/she is trying to run on his/her hardware). As good example touched above - the valuable third party updates for real modern hardware are not needed at all after you narrowed your request to using inside VM.

AFAIK on VMs the only thing you really need on modern hardware is TLB ivalidation patch for 98/ME plus speed patch for 95/98. All in one place - https://github.com/JHRobotics/patcher9x Literally nothing to talk.

For pure Windows 98SE Rus & Eng I maintain collection of raw QFE and some other updates with descriptions. And look at the BolenB sets for integration into distribution media. But it primarily depends on your hardware, of course.

I run 98 SE on Acer Aspire V3-571G and HP 250 G7 (84A6). Without some functionality, of course, but OTOH without errors.

Thank to @defuser testing: http://sweetlow.orgfree.com/download/a20alwon.zip 1. Fixed bug of simultaneous using /TESTMEM:ON and /A20CONTROL:OFF. Memory testing procedure relied on the state after /A20CONTROL:ON execution. 2. The code for the /A20CONTROL:OFF patch is made more generic and yet at the same time simplified. 3. Added 3+2+1 variant of patch. 4. The difference of variants -1 and 3 from /A20CONTROL:OFF on the hardware with the A20 Line Always On mentioned.

Intel Core2 Duo. The modern system built ... 15 years ago.

There are MANY protection errors possible in Windows 9x on modern hardware. It is far better to give exact message. SATA on modern laptop? Are you joking?

No, it is working as expected - XMS state is exactly equal to the state of real hardware. Always On to be precise. Ok, hardware is functioning as described above.

2_1 - as readme states. And it's the best till it's working, of course. Which of variants? You have to test 2 variants: 1. Original HIMEM.SYS (NOT in package). 2. 1_PS_2 Without any options except /v. And better with DOS=LOW in CONFIG.SYS (or without DOS too).