Re: [Freedos-user] XMS on a 8086

2025-07-31 Thread Liam Proven via Freedos-user
On Thu, 31 Jul 2025 at 02:40, tsiegel--- via Freedos-user wrote: > > The NEC V20 chip didn't have the whole 286 command set, but it > apparently had enough of it that *some* 286 specific programs would work. I never tried it myself, but the most famous example of this was the Windows 3.0 VGA driv

Re: [Freedos-user] XMS on a 8086

2025-07-31 Thread userbeitrag--- via Freedos-user
On 30th Aug 2025, 13:57, tsiegel--- via Freedos-user wrote: On 7/30/2025 6:41 AM, Ralf Quint via Freedos-user wrote: Sorry to disappoint you, but you can't run a "80826 specific programs on a V20". Wrong. Just to throw in one more "item": "8088 CPU is slightly slower than NEC V20, but

Re: [Freedos-user] XMS on a 8086

2025-07-31 Thread Christopher Evans via Freedos-user
Perhaps, scripts should be more prevalent instead of microcode and depend on is api layer interface calls. Be more inter compatible. .  To: "Ralf Quint via Freedos-user" ;Cc: "tsie...@softcon.com" ;Subject: [Freedos-user] XMS on a 8086;17:38, July 30, 2025, "tsiegel--- via Freedos-user" :On 7/30/20

Re: [Freedos-user] XMS on a 8086

2025-07-30 Thread tsiegel--- via Freedos-user
On 7/30/2025 6:41 AM, Ralf Quint via Freedos-user wrote: Sorry to disappoint you, but you can't run a "80826 specific programs on a V20". Wrong. I've done it, multiple times. More than likely, those programs didn't use any of the extended mode functionality available on a 286, but nonet

Re: [Freedos-user] XMS on a 8086

2025-07-29 Thread Ralf Quint via Freedos-user
On 7/29/2025 8:26 AM, tsiegel--- via Freedos-user wrote: And the fact that the V20/V30 chips understand (most of) the 80186 command set is because NEC decided to add those, just like the added 8080 emulation mode. Beside being a tad faster than the equivalent Intel chip, there is no function

Re: [Freedos-user] XMS on a 8086

2025-07-29 Thread tsiegel--- via Freedos-user
On 7/28/2025 6:55 PM, Ralf Quint via Freedos-user wrote: On 7/25/2025 2:17 PM, tsiegel--- via Freedos-user wrote: the 8088 chips (at least the nec V20 my xt machine had) could handle the 80186 command set, this allowed it to run (some, most?) 80286 programs.  I was able to run 286 specific ver

Re: [Freedos-user] XMS on a 8086

2025-07-29 Thread Ralf Quint via Freedos-user
On 7/28/2025 4:51 PM, Christopher Evans via Freedos-user wrote: You mod the quark occiliator to timings?, 42? ___ Freedos-user mailing list Freedos-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-user

Re: [Freedos-user] XMS on a 8086

2025-07-29 Thread Christopher Evans via Freedos-user
Quartz meant.  To: "Discussion and general questions about FreeDOS." ;Cc: Christopher Evans ;Subject: [Freedos-user] XMS on a 8086;17:16, July 28, 2025, "Christopher Evans via Freedos-user" :You mod the quark occiliator to timings?, To: "freedos-user@lists.sourceforge.net"

Re: [Freedos-user] XMS on a 8086

2025-07-28 Thread Christopher Evans via Freedos-user
You mod the quark occiliator to timings?, To: "freedos-user@lists.sourceforge.net" ;Cc: Ralf Quint ;Subject: [Freedos-user] XMS on a 8086;11:59, July 28, 2025, "Ralf Quint via Freedos-user" :On 7/25/2025 2:17 PM, tsiegel--- via Freedos-user wrote: 8088 isn't the same as 8086.  It's similar, but int

Re: [Freedos-user] XMS on a 8086

2025-07-28 Thread Ralf Quint via Freedos-user
On 7/25/2025 2:17 PM, tsiegel--- via Freedos-user wrote: 8088 isn't the same as 8086.  It's similar, but internally, the 8088 is different than the 8086, so that might influence the capabilities of the program.  You'll need to emulate a straight 8086 to get the full picture as to whether it ca

Re: [Freedos-user] XMS on a 8086

2025-07-26 Thread tom ehlert via Freedos-user
>> the 8088 chips (at least the nec V20 my xt machine had) > The NEC V20 was intended to run the 80186 instruction set, at somewhat higher > frequency. So in your strict definition > the NEC V20 is not a 8088, but a superset of it. additionally, the 80186/80188 and V20 would execute some inst

Re: [Freedos-user] XMS on a 8086

2025-07-26 Thread tom ehlert via Freedos-user
> 8088 isn't the same as 8086.  It's similar, but internally, the 8088 is > different than the 8086, Which is pretty much nonsense. The 8088 has an external 8 bit data bus, the 8086 has 16 bit. That's all the difference. Programs execute identical, except for performance. > so that might infl

Re: [Freedos-user] XMS on a 8086

2025-07-25 Thread tsiegel--- via Freedos-user
8088 isn't the same as 8086.  It's similar, but internally, the 8088 is different than the 8086, so that might influence the capabilities of the program.  You'll need to emulate a straight 8086 to get the full picture as to whether it can handle them or not. the 8088 chips (at least the nec V

Re: [Freedos-user] XMS on a 8086

2025-07-25 Thread Mateusz Viste via Freedos-user
Quick update on EXMS86: I have published v0.9.2 today. This version works around fragile EMS 4.0 implementations on some EMS cards. EXMS86 should be much more universal now. I have also confirmed that FreeCOM and its XMS-swap feature works on an 8086 (I tested this with 86box set to an emulati

Re: [Freedos-user] XMS on a 8086

2025-07-20 Thread Michael Brutman via Freedos-user
Ok, but to be very clear, 8086 class PCs do not have extended memory because they do not physically have the hardware. What you have implemented is a clever XMS API that uses EMS as the backing store. There is a difference between those two terms, w

Re: [Freedos-user] XMS on a 8086

2025-07-20 Thread Mateusz Viste via Freedos-user
On 20/07/2025 07:23, Michael Brutman wrote: How can an 8086 PC have a few megabytes of XMS memory?  By definition an 8086 PC only has 20 address lines, which gets you to 1MB.  There is no addressing of memory above 1MB available given that number of address lines. XMS (v2) is just a copying AP

Re: [Freedos-user] XMS on a 8086

2025-07-19 Thread E. C. Masloch via Freedos-user
Hi Eric, On at 2025-07-20 01:25:47 +0200, Eric Auer via Freedos-user wrote: > >Hi ECM, > >> I added XMS support (Extended Memory that isn't lockable, >> HMA with A20 always enabled, and UMBs) to 8086tiny. > >I do not understand. 8086 only has address lines A0 to A19, >how can XMS, HMA and A20 ex

Re: [Freedos-user] XMS on a 8086

2025-07-19 Thread Michael Brutman via Freedos-user
How can an 8086 PC have a few megabytes of XMS memory? By definition an 8086 PC only has 20 address lines, which gets you to 1MB. There is no addressing of memory above 1MB available given that number of address lines. ___ Freedos-user mailing list Free

Re: [Freedos-user] XMS on a 8086

2025-07-19 Thread Mateusz Viste via Freedos-user
Eric Auer wrote: >That makes perfect sense to me. FreeCOM is welcome to make >use of 286 optimizations given that PC which have sensible >amounts of XMS available must be PC which have 286+ CPU. FreeCOM does not need "sensible" amounts of XMS to swap. 150k is enough. Besides, an 8086 PC might v

Re: [Freedos-user] XMS on a 8086

2025-07-19 Thread Eric Auer via Freedos-user
Hi ECM, I added XMS support (Extended Memory that isn't lockable, HMA with A20 always enabled, and UMBs) to 8086tiny. I do not understand. 8086 only has address lines A0 to A19, how can XMS, HMA and A20 exist on 8086 hardware at all? In theory, you could implement something which copies aro

Re: [Freedos-user] XMS on a 8086

2025-07-19 Thread E. C. Masloch via Freedos-user
Hi, back a while ago I added XMS support (Extended Memory that isn't lockable, HMA with A20 always enabled, and UMBs) to 8086tiny. I found that at the time FreeCOM expected to run on a 186 or 286 when using XMS. (The 286 instructions unrelated to Protected Mode are all present on the 186 alread