Image Map Image Map
Page 37 of 41 FirstFirst ... 27333435363738394041 LastLast
Results 361 to 370 of 405

Thread: XTIDE Universal BIOS v2.0.0 beta testing thread

  1. #361

    Default

    Quote Originally Posted by Holmes View Post
    That's a problem, because I can't boot into 6.22 from any machine with a 5.25 floppy. Hmmm.
    How about if I give you a link to an image that you can copy to your PC and extract it right from there onto a floppy?
    If you're looking for DS/DD or DS/HD 3" or 5" floppy disks, PM me. I've got some new, used, and factory over-labeled disks for sale.

    There is a crack in everything. That's how the light gets in. -- Leonard Cohen
    ☞ Time flies like an arrow; fruit flies like a banana.

  2. #362
    Join Date
    Dec 2010
    Location
    Frederick, MD
    Posts
    259

    Default

    Quote Originally Posted by Stone View Post
    How about if I give you a link to an image that you can copy to your PC and extract it right from there onto a floppy?
    Thanks, but I managed to do it by finding a minimal dos 6.22 image and dskimage.exe for dos, transferring them over to the 5150, and writing the image directly from the 5150.

    Finally! The only problem is that the 6.22 version of xcopy still does NOT have the /h flag. Now why did I think it did have it?

  3. #363

    Default

    Quote Originally Posted by Holmes View Post
    Finally! The only problem is that the 6.22 version of xcopy still does NOT have the /h flag. Now why did I think it did have it?
    I dunno... I would have thought it did as well. But you are right -- it doesn't.

    Grab a copy of XXCOPY16 and use that. It's a third party program and is not DOS dependant.

    There's also a FreeDOS XCOPY that I've been using which has the /H switch.
    If you're looking for DS/DD or DS/HD 3" or 5" floppy disks, PM me. I've got some new, used, and factory over-labeled disks for sale.

    There is a crack in everything. That's how the light gets in. -- Leonard Cohen
    ☞ Time flies like an arrow; fruit flies like a banana.

  4. #364
    Join Date
    Dec 2010
    Location
    Frederick, MD
    Posts
    259

    Default

    Quote Originally Posted by Stone View Post
    I dunno... I would have thought it did as well. But you are right -- it doesn't.

    Grab a copy of XXCOPY16 and use that. It's a third party program and is not DOS dependant.

    There's also a FreeDOS XCOPY that I've been using which has the /H switch.
    XXCOPY16 does look comprehensive. Dizzying array of options. It does seem to have a switch set called /clone which seems to be a good option for exactly what I want to do.

    Freedos looks like I'd need to install it, so I think I'll try xxcopy16.

    Thanks!

  5. #365

    Default

    Quote Originally Posted by Holmes View Post
    XXCOPY16 does look comprehensive. Dizzying array of options. It does seem to have a switch set called /clone which seems to be a good option for exactly what I want to do.

    Freedos looks like I'd need to install it, so I think I'll try xxcopy16.

    Thanks!
    XXCOPY16 also has the /H switch in addition to the /clone switch!

    The XCOPY I have from Freedos does not need to be installed.

    If you're looking for DS/DD or DS/HD 3" or 5" floppy disks, PM me. I've got some new, used, and factory over-labeled disks for sale.

    There is a crack in everything. That's how the light gets in. -- Leonard Cohen
    ☞ Time flies like an arrow; fruit flies like a banana.

  6. #366
    Join Date
    Dec 2010
    Location
    Frederick, MD
    Posts
    259

    Default

    Quote Originally Posted by Stone View Post
    XXCOPY16 also has the /H switch in addition to the /clone switch!

    The XCOPY I have from Freedos does not need to be installed.

    Thanks for the link! I'll keep that in my PC box of tools for the future.

    I used XXCOPY16, and it did what I needed, and seems to have worked.

    Thanks again for all your help.

  7. #367
    Join Date
    Mar 2011
    Location
    Atlanta, GA, USA
    Posts
    1,120

    Default ATA emulator problems

    Odd request. I'm testing my IDE emulator with a Glitchworks XT-IDEv3 and Universal BIOS 2.0.0beta3 on a 5150 PC w/ 64KB. After implementing a few commands I didn't expect to be issued, I'm not getting correct data in BIOS calls. However I can fall into ROM BASIC and perform manual inp/outs and see that READ_MULTIPLE commands are working, the correct statuses are being set, and I'm getting the correct data in the latch registers. So I'm stumped...

    The BIOS doesn't print any ID string - almost like it saw nothing for the ID sector. But it does detect master and slave presence as I turn them on/off in the emulator. And of course I get boot sector not found and it fails to boot drive c. But I can issue an ID command in BASIC and see a well formed ID sector just fine.

    Quick summary of what my emulator does:

    1) When the command register is written, all the error bits are cleared, the busy bit is set, and the DRQ bit is cleared. All data transfer FIFOs are also reset to empty.
    2) Once the back-end of the emulator services the command and it's arguments, it clears the BSY bit, ensures DSC is set, DRDY is set, all error bits are cleared.. but leaves DRQ off for now.
    3) Once enough data has built up in the transfer FIFO such that a 256 word transfer can be performed at any speed (guaranteed no underflow), DRQ is set until 256 words actually get transferred.

    This all pretty standard ATA spec stuff. My emulator does not support IRQ mode at the moment.

    This is a quickly thrown together test setup - hence the 64KB 5150. I don't have a workable boot disk which is why I'm relying on ROM BASIC for testing. I can try to throw something more substantial together if there are no quick suggestions here.

    Thanks,

    -Alan
    "Good engineers keep thick authoritative books on their shelf. Not for their own reference, but to throw at people who ask stupid questions; hoping a small fragment of knowledge will osmotically transfer with each cranial impact." - Me

  8. #368
    Join Date
    Mar 2011
    Location
    Atlanta, GA, USA
    Posts
    1,120

    Default

    I believe I have a theory, but cannot test until tonight. I think the XT-IDEv3 is violating the ATA setup and hold times wrt the upper/lower byte latches. Because the latch enables are being shoved through so much combinatorial logic on the board -> '688 address comparitor for IOSEL, further qualification of A0=0, A1=0, A2=0, the '138 mux selection for control strobes, and the inverters to right the logic, etc... and the IDE_RD signal is just a '245 buffered ISA_RD, my emulator is dropping the data presentation on the IDE data lines before the XT-IDEv3 has a chance to latch it in - on the upper byte.

    I'll try delaying the data presentation some high number of internal clocks past strobe de-assertion and see if that helps. If the theory holds true however, the XT-IDEv3 is pretty flawed and compatibility problems will only get worse with newer generation PATA parts. Good thing not many are made any longer!
    "Good engineers keep thick authoritative books on their shelf. Not for their own reference, but to throw at people who ask stupid questions; hoping a small fragment of knowledge will osmotically transfer with each cranial impact." - Me

  9. #369
    Join Date
    Mar 2011
    Location
    Atlanta, GA, USA
    Posts
    1,120

    Default

    Finally fixed it. It was a combination of the above and what I consider a serious design flaw in XT-IDE.

    1) I had to add an extended hold delay that would persist the 16-bit IDE data while keeping all the data buffers enabled a configurable amount of time after the read strobe was de-asserted by XT-IDE.

    2) For some reason the original XT-IDE designer decided to decode the high/low extended byte latch at the same address as the IDE device's CS1 address 0. While there is no ATA register mapped there, I had initially mapped my emulator's console UART data register (Raspberry Pi backend) to that address. The high byte latch and the IDE device were both driving the IDE data lines. While I suspect most IDE devices may either float the data bus during CS1:0 accesses, some probably do not. And it's reasonable to think many wont.

    I suspect one or both of thee problems are causing most of the drive compatibility problems with XT-IDE cards of any version. At least my 5150 is booting DOS 2.1 now! yay!
    "Good engineers keep thick authoritative books on their shelf. Not for their own reference, but to throw at people who ask stupid questions; hoping a small fragment of knowledge will osmotically transfer with each cranial impact." - Me

  10. #370

    Default

    Quote Originally Posted by eeguru View Post
    Odd request. I'm testing my IDE emulator with a Glitchworks XT-IDEv3 and Universal BIOS 2.0.0beta3 on a 5150 PC w/ 64KB. After implementing a few commands I didn't expect to be issued, I'm not getting correct data in BIOS calls. However I can fall into ROM BASIC and perform manual inp/outs and see that READ_MULTIPLE commands are working, the correct statuses are being set, and I'm getting the correct data in the latch registers. So I'm stumped...

    The BIOS doesn't print any ID string - almost like it saw nothing for the ID sector. But it does detect master and slave presence as I turn them on/off in the emulator. And of course I get boot sector not found and it fails to boot drive c. But I can issue an ID command in BASIC and see a well formed ID sector just fine.
    I recommend using the latest revision available from www.xtideuniversalbios.org/binaries for your testing.

    Beta 3 will ignore drives that don't pass the ATA-ID validation check. This check is simply an attempt to avoid data corruption on the drive if there's a problem communicating with it (due to bad cabling, soldering etc). However, it was discovered that a lot of perfectly good Compact Flash cards would fail the check and later on two different Western Digital Caviar models were found to also be incompatible with the ATA-ID validation so it was completely disabled in revision 580. (The code is still in there though and it can be re-enabled by removing the NO_ATAID_VALIDATION define from the makefile before building.)

    Quote Originally Posted by eeguru View Post
    If the theory holds true however, the XT-IDEv3 is pretty flawed and compatibility problems will only get worse with newer generation PATA parts. Good thing not many are made any longer!
    So is this a problem with the XT-IDE v3 only or has it been confirmed to be a problem with all three versions of the XT-IDE card? I've seen comments on here indicating a timing problem with the XT-IDE v3 card but nothing really conclusive.
    Looking for a cache card for the "ICL ErgoPRO C4/66d V"

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •