Re: My MSX project + protect mode
On Sat, 28 Nov 1998, Jun-Sung Kim wrote: > Date: Sat, 28 Nov 1998 10:33:42 +0900 (KST) > From: Jun-Sung Kim <[EMAIL PROTECTED]> > Reply-To: [EMAIL PROTECTED] > To: [EMAIL PROTECTED] > Subject: Re: My MSX project + protect mode > > > Hi, > > Lately about the protected mode... > > If you want to implement it, minimize it. > The reason that I like MSX, is the full controllability. MSX is a > PERSONAL Computer, so all results are responsible to the user. > > Well, it may be also good to prevent the important system area > being crushed in another point of view. Then I suggest that > just some memory space can be protected optionally. In AHDL style :), > > mem_protect = ( A[31] == b"1" ) and not(super); > > The simplest way! ( I like always "simple" :) > > By the way, did you fix the memory usage(physical/logical memory > configurations) and the slot pin assignment? > > - Jun. If you want to make a new memory-configuration, and it should be multitasking (please, make it linux-like and not windows-like), I have a tip for you that most pc-programmers will not agree with: Make it possible to use segments. What I mean is you can set some register outside the cpu on a value rg and if an address ad is asked by the cpu, the address that is actually read out is ad+mf*rg, where mf is a multiplying factor, prefferibly smaller than 64 (use a power of 2 of course, so you don't actually have to multiply). This makes it possible to load a program into just about any address, and start it from wherever you want the origin to be. If you would use standard msx-memory configuration for multitasking, you would have to use at least 16kB for every task, no matter how small. It's not too hard to implement, but it makes life much easier for programmers. (I don't know why pc-programmers always say segments are the worst thing ever invented...) Bye, shevek MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
> Make it possible to use segments. What I mean is you can set some register > outside the cpu on a value rg and if an address ad is asked by the cpu, > the address that is actually read out is ad+mf*rg, where mf is a > multiplying factor, prefferibly smaller than 64 (use a power of 2 of > course, so you don't actually have to multiply). Why a small value like 64? A Z380 based system will probably have a memory size in the order of megabytes, so a value like 64K will be acceptable as well. Ofcourse, if a value of 1 is possible, that would be ideal, but maybe that will make hardware implementation too complex. > It's not too hard to implement, but it makes life much > easier for programmers. Implementation may be simple, but it may have drastic impact on de speed of the system. Memory accesses are very important for how fast a system runs (every instruction must be fetched from memory). If the adding of a segment offset takes additional clockticks for every memory access, it will seriously slow down the system. There are other ways of solving problems with relocation: - Using relative addressing most of the time. For example JR and DJNZ, but also using IX and IY registers. Z380 has stackpointer relative addressing, which makes building compilers much easier. - Filling in the right absolute addresses after loading an executable. If a standardized mechanism is used for this, it is only a small burden. > (I don't know why pc-programmers always say > segments are the worst thing ever invented...) The decision of 16 bytes increments for segments was wrong. That decision is the reason x86 real mode can only address 1024K, so indirectly it is responsible for the hated 640K boundary. Bye, Maarten MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
Shevek wrote: >If you want to make a new memory-configuration, and it should be >multitasking (please, make it linux-like and not windows-like), I agree with this. >I have a >tip for you that most pc-programmers will not agree with: >Make it possible to use segments. N!! Not segments!!! :( >(I don't know why pc-programmers always say >segments are the worst thing ever invented...) Well, maybe because they KNOW WHAT THEY'RE TALKING ABOUT!? Segments are about the crappiest thing you can imagine... Z380 allows fully relocatable programs, it has JR and CALR (CALL relative!) that can jump 8MB in either direction! Greetz, Patriek ,--. ,---. ,--. Homepage: \"To make a mistake is | '--.| __ \ \__/ www.xs4all.nl/~newimage/ / human, but to really | __|| | | | ,--. E-mail: \ fuck things up, you | | | | | | | | [EMAIL PROTECTED]/ need a computer." | '---' | | '--' | \- Glenn Scott, \_| || The New Image -since 1991-/ Secret Agent W7 MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
Hi All I do not answer on Your letters. Because they have no any questions. Indeed I attentively read them. Thank you very mutch For help :-) Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
>I have a slightly different idea. Allow write protect option for a part of >the RAM. Just like turbo R BIOS in DRAM, or ESE-SCC in ROM mode. > Can You Write me about details? Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project
Hi > There are several I/O controller chips for PC. I consider to use >PC87309 of National Semiconductor. It includes keyboard/mouse >controller, floppy disk controller, parallel and serial ports. > How about hardware compatibility. PC87.. have own i/o adress space. KBD is it problem too. RS232 must +-12V. ( I like more RS422 ) Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
Hi, Lately about the protected mode... If you want to implement it, minimize it. The reason that I like MSX, is the full controllability. MSX is a PERSONAL Computer, so all results are responsible to the user. Well, it may be also good to prevent the important system area being crushed in another point of view. Then I suggest that just some memory space can be protected optionally. In AHDL style :), mem_protect = ( A[31] == b"1" ) and not(super); The simplest way! ( I like always "simple" :) By the way, did you fix the memory usage(physical/logical memory configurations) and the slot pin assignment? - Jun. MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
At 10:35 AM 11/25/98 +0600, you wrote: >As possible do protected mode? > >Why? For raising system reliability. > >Processor Must have 2 modes system and user. Bit "super" >In the field of memories is conducted place which is only read in mode >"user". >If processor causes function of system that super=on. Access To registers of >governing protection of memory possible in this mode only. >Preceding mode returns When leaving I have a slightly different idea. Allow write protect option for a part of the RAM. Just like turbo R BIOS in DRAM, or ESE-SCC in ROM mode. Advantages: - not only system code can be protected, but all code (incl user code) - probably simpler to implement, because mode switching is not necessary for every system call Disadvantages: - system data is not protected >Be >also possibility to lengthen mapper before 16 bits of command Z80 will be >able to change only younger byte, Z380 both using OUTW If a programmer wants to use more than 4MB of memory (limit of 8-bit mapper port), I don't think he'll want anything else than linear memory. In other words, 16-bit mapper port has a low practical value. Bye, Maarten MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
Hi, > If you want to implement it, minimize it. >The reason that I like MSX, is the full controllability. MSX is a >PERSONAL Computer, so all results are responsible to the user. I want to make protect mode not for user. It will be for crash-programm. User can free to change "super" bit. > > > mem_protect = ( A[31] == b"1" ) and not(super); > If will be crash a programm. She can write to area Where A31 == 0. I wilI be to make it ease. > > By the way, did you fix the memory usage(physical/logical memory >configurations) and the slot pin assignment? Not at all. More MSX users offer using PCI. I can not to dare. It will be dearly. On the other hand this minimum current standard. And can be we emergency its use. Presently we do mother-board but we will come to do and graphic card. v9990 this not that needed me. But I sees difficulties in use PC-graph-card. Inconsistency with each other one of them. On this PCI not it is necessary Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project
Hi MSXers, > > > Do you mean that 00-3F can be used both for linear addressing > >and memory mapping (MSX2 standard)? > > > Yes > > Do You think on other? > I, myself, thought to make two separate memories. When access below 64KB space, memory module A(4MB, memory mapper, MSX2 standard) is accessed. Otherwise, memory module B(16MB or more for Z380) is accessed. In other words, same physical memory location is not accessed via two different address. My approach wastes 64KB of memory module B. But it's possible to use the space as usefull system memory. Just one suggestion. > > > How about to use super I/O controller chip for PC? > > > > Where to take a specification and chips themselves? > That this gives? > There are several I/O controller chips for PC. I consider to use PC87309 of National Semiconductor. It includes keyboard/mouse controller, floppy disk controller, parallel and serial ports. > Yes This is one variant. The Second variant when Z8 itself as PPI, it is > more better. > OK. The second one is better. Complete compatibility to both side, PPI and KBD. Greetings, - Jun. MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
On Tue, 1 Dec 1998, Patriek Lesparre wrote: > Date: Tue, 01 Dec 1998 04:54:43 +0100 > From: Patriek Lesparre <[EMAIL PROTECTED]> > Reply-To: [EMAIL PROTECTED] > To: [EMAIL PROTECTED] > Subject: Re: My MSX project + protect mode > > Shevek wrote: > >If you want to make a new memory-configuration, and it should be > >multitasking (please, make it linux-like and not windows-like), > > I agree with this. > > >I have a > >tip for you that most pc-programmers will not agree with: > >Make it possible to use segments. > > N!! Not segments!!! :( > > >(I don't know why pc-programmers always say > >segments are the worst thing ever invented...) > > Well, maybe because they KNOW WHAT THEY'RE TALKING ABOUT!? Segments are > about the crappiest thing you can imagine... I still don't see why... > Z380 allows fully relocatable programs, it has JR and CALR (CALL relative!) > that can jump 8MB in either direction! In that case, I agree, you don't need segments. > Greetz, > > Patriek Bye, shevek MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project + protect mode
At 08:46 PM 11/26/98 +0600, you wrote: >>I have a slightly different idea. Allow write protect option for a part of >>the RAM. Just like turbo R BIOS in DRAM, or ESE-SCC in ROM mode. > >Can You Write me about details? Using write protect: Let's say you have a Z380 system with 4MB RAM, at addresses 00-3F. System software code is 256K in size, system software data is another 256K. The code is placed at 00-03, the data at 04-07. You can choose to write protect only 00-03, giving low protection but less overhead, or write protect entire 00-07 area, giving better protection but demanding the system software to turn off write protect before modifying its data and turning write protect on afterwards. Implementing write protect: You could use a address mask, allowing only sizes that are a power of 2 to be write protected (..., 01, 02, 04, ...). I think you suggested something like that in your previous message, but I'm not sure I understood all of it. A more flexible way is to make segments of a fixed size, and decide per segment wether it can be written or not. Segment size could be 64K for example. This would be very flexible: - you can protect variable sizes of memory (not just powers of 2) - in a multitasking system, you could write protect memory of idle tasks But the costs are also high: you have to do a table lookup for every memory access. Unless you can implement this very fast in hardware, it would bring down system performance a lot. How to handle an illegal memory write: I see two options: - Ignore the write attempt. This can be done by not sending the memory IC the write enable signal. Advantage: simple to implement. Disadvantage: the user won't know a program is behaving badly. - Generate some kind of exception. For example, generate an interrupt and use a certain flag to incidate a write violation. The operating system can act upon this, for example by shutting down the offending program. If you have more questions, please ask them. Bye, Maarten MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
My MSX project + protect mode
Hi
As possible do protected mode?
Why? For raising system reliability.
Processor Must have 2 modes system and user. Bit "super"
In the field of memories is conducted place which is only read in mode
"user".
If processor causes function of system that super=on. Access To registers of
governing protection of memory possible in this mode only.
Preceding mode returns When leaving
That this for registers?
For example:
Mapper Mask Register MMR[7..0] Kernel will be able to select necessary
amount of pages for MSX & DOS2
MMR[].d=d[7..0];
MMR[].clk=/IOWR;
MMR[].ena=(A[7..0]== { adres of MMR } );
MMR[].reset=reset;
IF super THEN
mask[7..0]=B""
ELSE
mask[7..0]=MMR[7..0].q
END IF;
Cycle write to mapper
Mapper[7..0]=d[7..0] and not(mask[7..0]);
Cycle read to mapper
d[7..0]=Mapper[7..0] or mask[7..0];
Example two: Kernel can protect upper area memories by permit 16 kb ( pseudo
ROM )
MemoryProtectRegister MPR[7..0]
LoMem=(A[23..16]==0)
IF LoMem THEN
IMA[21..14]=mapper[7..0];-- IMA is internel adress
ELSE
IMA[21..14]=A[21..14];
END IF;
mem_protect=( IMA[21..14]=> MPR[7..0].q) and not(super);
Example3: Protect invalid -opcode vector
mem_protect=(A[23..2]==0) & not(super)
If think possible to do anything else. Main that this was not complex. Be
also possibility to lengthen mapper before 16 bits of command Z80 will be
able to change only younger byte, Z380 both using OUTW
But that will say you ? :)
Valery
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
My MSX project - Protect mode
Hi
As possible do protected mode?
Why? For raising system reliability.
Processor Must have 2 modes system and user. Bit "super"
In the field of memories is conducted place which is only read in mode
"user".
If processor causes function of system that super=on. Access To registers of
governing protection of memory possible in this mode only.
Preceding mode returns When leaving
That this for registers?
For example:
Mapper Mask Register MMR[7..0] Kernel will be able to select necessary
amount of pages for MSX & DOS2
MMR[].d=d[7..0];
MMR[].clk=/IOWR;
MMR[].ena=(A[7..0]== { adres of MMR } );
MMR[].reset=reset;
IF super THEN
mask[7..0]=B""
ELSE
mask[7..0]=MMR[7..0].q
END IF;
Cycle write to mapper
Mapper[7..0]=d[7..0] and not(mask[7..0]);
Cycle read to mapper
d[7..0]=Mapper[7..0] or mask[7..0];
Example two: Kernel can protect upper area memories by permit 16 kb ( pseudo
ROM )
MemoryProtectRegister MPR[7..0]
LoMem=(A[23..16]==0)
IF LoMem THEN
IMA[21..14]=mapper[7..0];-- IMA is internel adress
ELSE
IMA[21..14]=A[21..14];
END IF;
mem_protect=( IMA[21..14]=> MPR[7..0].q) and not(super);
Example3: Protect invalid -opcode vector
mem_protect=(A[23..2]==0) & not(super)
If think possible to do anything else. Main that this was not complex. Be
also possibility to lengthen mapper before 16 bits of command Z80 will be
able to change only younger byte, Z380 both using OUTW
But that will say you ? :)
Valery
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project
Hi, > > Hi > I wrote I want build MSX-3 ( non ASCII standart ) > I wrote FPGA-Mem memory controller for my MSX3 computer. It take > EPM7128SLC84. I writing other FPGA-IO now. > > FPGA-Mem specification: > Using one or two 16Mbit 60-nc DRAM 1MbX16Bits > > Z380 Memory map > > 00-00MSX MEMORY > 01-3FDRAM (2MB x 16Bit) > 40-7FEXTENDED MSX SLOT 0 > 80-BFEXTENDED MSX SLOT2 > C0-FFINTERNAL MSX SLOT ( FOR FLASH ROM DISK) > What is the EXTENDED MSX SLOT and what's the difference between 40- and 80- ? Also why INTERNAL MSX SLOT is needed? Hmm.. it is possible for me to misunderstand your SLOT concept. > MSX MEMORY > -0003ROM-BIOS -- for invalid opcode exeption > 0004-3FFF CS0 > 4000-7FFF CS1 > 8000-BFFF CS2 > C000- CS3 > I think that the first 64KB should be same as MSX standard. So we need not to discuss this address space. Right? > CS[0..3] tru via port A8 & mapper. Y can using mapper for acces > 00-3F DRAM > Do you mean that 00-3F can be used both for linear addressing and memory mapping (MSX2 standard)? > SLOT [A8] MAP control FLASH ROM 128KB > > SLT0 CS0 BIOS > SLT0 CS1BASIC > SLT1 CS0XBIOS > What's XBIOS? New BIOS for the new standard? > SLT1 CS1BDOS > SLT2 CS1DOS2 ( 64KB controll 4104H ) ^^^ I'm sorry? What does it mean? > SLT3 CS[0..3] MAPPER DRAMM ACCES > > I want place into the FPGA-Mem MEMORY and I/O acces-NMI trapper. > > Other FPGA will has EPP , SLOTS ARBITR, I/O decoder for VDP, IDE, 85C30 > serial, CMOS, AY and KBD . > How about to use super I/O controller chip for PC? > KBD controller will use Z8 microcontroller for PC-KBD to MSX-KBD and > MS-mouse to MSX-mouse convertions. > The problem of KBD is that MSX does not have KBD connector. Also MSX controls KBD through PPI. How do you connect the converter to PPI? > Extended MSX slot it is my last problemm. We want compatibility. And We can > take 62pins ISA connector. Get out 51,52 pins and insert wall. > > 1 2 > > | | > | | > > > > | | > | | 49,50 > wall > | | 53,54 Extended MSX > | | > | | 61,62 > > > Extended MSX must has, > A[16..23],D[8..15],/BLEN,/BHEN,/IORD,/IOWR,/MRD,/MWR,/MSIZE,/BREQ,/BACK (It > is 25 pins) .But Extended part has 10 pins. MSX slot has two-spare , /rfsh > ... it is all. May be We can use SW1,SW2,-12, +12, sound in e.t.c > Make another different slot for new standard. PC has several types of slots (XT ISA, AT ISA, PCI, AGP, ... etc. ) > We can take PCI or PCMCIA conector or make PCMCIA slot. I do not know yet. > We did want make PCI slot but It will hi-cost. > In my opinion, MSX should be simple. Should such slots be included? To tell the truth, I hate PCI protocol. ;) Ok. Keep you project and I think we need a lot of discusstions to prevent creating diffrent MSX standards. Greetings, - Jun. MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
Re: My MSX project
Hi, > What is the EXTENDED MSX SLOT and what's the difference between 40- >and 80- ? > It is different signals /SLTSL0 and /SLTSL1 ( as beside MSX ) connected to different connectors. To avoid decoder of address on affiliated card > Also why INTERNAL MSX SLOT is needed? That which be divorced on maternal charge for instance FLASH ROM 1 or 2 or 4MB Port À8 does not let to handle such sizes. Extended slots powerfully slow a functioning (working) a system. > I think that the first 64KB should be same as MSX standard. So >we need not to discuss this address space. Right? Except cells -0004 and extended slot. Cell 0-4 needed for wrong-instruction trap. I want that, it will be available always. > Do you mean that 00-3F can be used both for linear addressing >and memory mapping (MSX2 standard)? Yes Do You think on other? > What's XBIOS? New BIOS for the new standard? > No It is MSX2 extended BIOS. >> SLT1 CS1BDOS >> SLT2 CS1DOS2 ( 64KB controll 4104H ) > ^^^ >I'm sorry? What does it mean? > DOS2 take 64 KB ROM. But All four 16KB pages are connected from address 4000 This controls a register 4104. Anyway that DOS2 which beside me. > How about to use super I/O controller chip for PC? > Where to take a specification and chips themselves? That this gives? > The problem of KBD is that MSX does not have KBD connector. Also >MSX controls KBD through PPI. How do you connect the converter to >PPI? Yes This is one variant. The Second variant when Z8 itself as PPI, it is more better. > Make another different slot for new standard. PC has several types of >slots (XT ISA, AT ISA, PCI, AGP, ... etc. ) Certainly. But I wants to take a good concept. Beside MSX slot it was. "Cheap controllers" I wants its save and add " Upgrade cpu via slot connector" > In my opinion, MSX should be simple. Should such slots be included? >To tell the truth, I hate PCI protocol. ;) I too. Presently I consider;examin variant amba-bus ( ARM ) It has a separate address and data. ( vs PCI ) But fast the whole will that that its > Ok. Keep you project and I think we need a lot of discusstions to prevent >creating diffrent MSX standards. I this wanted :) Soon open FTP for my lifelengths, all will be it is available gratis. Presently goes preliminary part a writing FPGA. I use not expensive from ALTERA that general prime cost was not more than 100$ Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
My MSX project & hardware MSX mail-list
Hi, Who know about MSX-hardware mail list?. If such be There is in view of hardware development. It is Correct have on slot-connector the pins /TREFA,/TREFC,/TREFR ? /TREFA Timing Reference A (output, active Low, tri-state). This timing reference signal goes Low at the end of T2 and returns High at the end of T4 during a memory read, memory write or refresh transaction. It can be used to control the address multiplexer for a DRAM interface or as the /RAS signal at higher processor clock rates. /TREFC Timing Reference C (output, active Low, tri-state). This timing reference signal goes Low at the end of T3 and returns High at the end of T4 during a memory read, memory write or refresh transaction. It can be used as the /CAS signal for DRAM accesses. /TREFR Timing Reference R (output, active Low, tri-state). This timing reference signal goes Low at the end of T1 and returns High at the end of T4 during a memory read, memory write or refresh transaction. It can be used as the /RAS signal for DRAM accesses. Wanted their use. But in the mode DMA affiliated charge must their form. This complicates its contents. /UMCS Upper Memory Chip Select (output, active Low, tri-state). This signal is activated during a memory read, memory write, or optionally a refresh transaction when accessing the highest portion of the linear address space within the first 16 Mbytes, but only if this chip select function is enabled. /LMCS Low Memory Chip Select (output, active Low, tri-state). This signal is activated during a memory read or memory write transaction when accessing the lower por-tion of the linear address space within the first 16 Mbytes, but only if this chip select function is enabled. /MCS3-/MCS0 Mid-range Memory Chip Selects (output, active Low, tri-state). These signals are individually active during memory read or write transactions when accessing the mid-range portions of the linear address space within the first 16 Mbytes. These signals can be individually enabled or disabled. Too the most. Why Zilog did not do a possibility of shaping these signals in DMA mode. Z380 in mode of seizure of buses could read address and form signals /XMCSX :) Bye Valery MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
My MSX project
Hi
I wrote I want build MSX-3 ( non ASCII standart )
I wrote FPGA-Mem memory controller for my MSX3 computer. It take
EPM7128SLC84. I writing other FPGA-IO now.
FPGA-Mem specification:
Using one or two 16Mbit 60-nc DRAM 1MbX16Bits
Z380 Memory map
00-00MSX MEMORY
01-3FDRAM (2MB x 16Bit)
40-7FEXTENDED MSX SLOT 0
80-BFEXTENDED MSX SLOT2
C0-FFINTERNAL MSX SLOT ( FOR FLASH ROM DISK)
MSX MEMORY
-0003ROM-BIOS -- for invalid opcode exeption
0004-3FFF CS0
4000-7FFF CS1
8000-BFFF CS2
C000- CS3
CS[0..3] tru via port A8 & mapper. Y can using mapper for acces
00-3F DRAM
SLOT [A8] MAP control FLASH ROM 128KB
SLT0 CS0 BIOS
SLT0 CS1BASIC
SLT1 CS0XBIOS
SLT1 CS1BDOS
SLT2 CS1DOS2 ( 64KB controll 4104H )
SLT3 CS[0..3] MAPPER DRAMM ACCES
I want place into the FPGA-Mem MEMORY and I/O acces-NMI trapper.
Other FPGA will has EPP , SLOTS ARBITR, I/O decoder for VDP, IDE, 85C30
serial, CMOS, AY and KBD .
KBD controller will use Z8 microcontroller for PC-KBD to MSX-KBD and
MS-mouse to MSX-mouse convertions.
Extended MSX slot it is my last problemm. We want compatibility. And We can
take 62pins ISA connector. Get out 51,52 pins and insert wall.
1 2
| |
| |
| |
| | 49,50
wall
| | 53,54 Extended MSX
| |
| | 61,62
Extended MSX must has,
A[16..23],D[8..15],/BLEN,/BHEN,/IORD,/IOWR,/MRD,/MWR,/MSIZE,/BREQ,/BACK (It
is 25 pins) .But Extended part has 10 pins. MSX slot has two-spare , /rfsh
... it is all. May be We can use SW1,SW2,-12, +12, sound in e.t.c
We can take PCI or PCMCIA conector or make PCMCIA slot. I do not know yet.
We did want make PCI slot but It will hi-cost.
Bye
Valery
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
RE: My MSX project + protect mode
-Mensaje original-
De: [EMAIL PROTECTED] [SMTP:[EMAIL PROTECTED]]
Enviado el: lunes 30 de noviembre de 1998 12:53
Para: [EMAIL PROTECTED]
Asunto: Re: My MSX project + protect mode
> Make it possible to use segments. What I mean is you can set some register
> outside the cpu on a value rg and if an address ad is asked by the cpu,
> the address that is actually read out is ad+mf*rg, where mf is a
> multiplying factor, prefferibly smaller than 64 (use a power of 2 of
> course, so you don't actually have to multiply).
Why a small value like 64? A Z380 based system will probably have a
memory size in the order of megabytes, so a value like 64K will be
acceptable as well.
Ofcourse, if a value of 1 is possible, that would be ideal, but maybe
that will make hardware implementation too complex.
[...]
> (I don't know why pc-programmers always say
> segments are the worst thing ever invented...)
The decision of 16 bytes increments for segments was wrong. That
decision is the reason x86 real mode can only address 1024K, so
indirectly it is responsible for the hated 640K boundary.
Bye,
Maarten
[Isidoro González Márquez] I would add that the wrong decision about
8086 PC segments, which made
them so
hated by everybody, was their
fixed length
(64K) and the absence (lack)
of other mecha-
nisms associated to modern
segmentation
(memory protection, read-only
possibility,
etc.).
[Isidoro González Márquez] One of the worst aspects of Z80 is the lack of
the possibility of addressing
relative to
PC. That would make programs
capable of running
on whatever address. 6809 was
much better
in this aspect. The right
solution is al-
ready invented: virtual
memory. All programs
are written as if they were
running on the
same address and the hardware
makes the
translation "on the fly".
Gyermo.
begin 600 WINMAIL.DAT
M>)\^(C<+`0:0" `$```!``$``0>0!@`(Y 0```#H``$(@ <`
M& ```$E032Y-:6-R;W-O9G0@36%I;"Y.;W1E`#$(`0V ! `"`@`"``$$
MD 8`: $```$0`P``, (+``\.``(!_P\!-P``
M``"!*Q^DOJ,0&9UN`-T!#U0"`&US>$!S=&%C:RYN; !33510`&US>$!S
M=&%C:RYN; ``'@`", $%4TU44 `>``,P`0T```!M`/9?`0T```!M`' `
M`0```"(```!213H@37D@35-8('!R;VIE8W0@*R!P`!X,`04`
M``!33510`!X`'PP!% ```&=Y97)M;T!G=6=U+G5S86PN97,``P`&
M$%(;`J8#``<0JP4``!X`"! !90```"TM+2TM345.4T%*14]224=)3D%,
M+2TM+2U$13I-2U1(55523D5 4U151%1514Y,4TU44#I-2U1(55523D5 4U15
M1%1514Y,14Y624%$3T5,.DQ53D53,S!$14Y/5DE%34)2140``@$)$ $`
M``"2!@``C@8``-8+``!,6D9U4,V(@1&4Z#(-B`1%P
M32Y++G0N2 )U"'!N94!S='4F9"!@"E N;@,@6U.P3510.B _(45=%]41'V!%
M;G8',&1O('QE;!^G"D @T 0@#T @T0$`(&YO)&!E!M )<"$E\C$Y.3@F\#(Z
M5#4S([=0"L!A'Z=M;'-X(/$`T&[EMAIL PROTECTED]=)S)8!T;Q^G4A^0!=!R>0716" 9
MD1LP&> @_BLKXAG"*/ $<1P_'4H8]$L7;PJ /@7086L;0&GI!4!P;P00:0)@
M&T JD-0@=1% (!% 9P> `C"P0A@(&/_`Y$1
M03) `W ;0 EP&Y A`.\$D#!6"& RL&DF`1(@&T"'#? T``(@(&$@=@= 2PI0
M-.!G-Q!N9#$09O,WP3<09&0)< 01)( SHFQA!U?0= ;"N0"7 X\37Q.1-D$"MM9BHWH"P@=Q\V@":A/: Y$S!6
M;75LWG0%(#R "X WL&8\(06POSWP$U !$09Q`F KD',`P \\,QZP.: Y4#F! !&!R04%I>G\Q`0.@-G(%L 2!0R('
M@&=U`:!Y$\+4"9P,I$S$&DV\2J0)*![!:!8\GA'%A#C_&G \L5?A!'$T$P(@/($X=G@Q,#)0
M`$2#%]0+@&3^:0EP&> \@3$A,[$XH3%0?P"!,:)FDC9R,P$Y@4(P,%=0$ ;@
M)8!D"L!Y1QM"_GE$K]R8G,>L&I!4$,R1SWR#>!H_U:1-D1*L#2@>A]_7V]E.;'W8H)6
MP 1P>3WQ4<$V<6U0^4 0:7@Y@3&P#R 2('Z/MX0O0E!/\2DWPSKC8A% ^V7P
M&T H"V I<(7 0S$2$?\1T0> $0$<)H1/>YD#`$%@_SDR-*!DL&^#1E($=#)5
M67/_B$^-OT)03(49E5F1/? \LOXM:U,Q5 ,0,2"!P(SODH_[7. 14&-C<%J(
M&7ASCW2?>W6O=KM/(-!#(F&(.5!P]QG1!"!#,5I)T6E5AO)#=_^2GWN*-G*0
M^4,B.'4
