Is it possible to boot OpenVMS from an IDE disk on an ES40? - DEC

This is a discussion on Is it possible to boot OpenVMS from an IDE disk on an ES40? - DEC ; Hi, I'm running into a problem with an emulator for the ES40 that I'm working on. I get the following error: jumping to bootstrap code %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910 I'm ...

+ Reply to Thread
Page 1 of 3 1 2 3 LastLast
Results 1 to 20 of 41

Thread: Is it possible to boot OpenVMS from an IDE disk on an ES40?

  1. Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Hi,

    I'm running into a problem with an emulator for the ES40 that I'm
    working on. I get the following error:

    jumping to bootstrap code
    %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    I'm trying to boot from an emulated IDE disk (not CD-ROM) because
    emulating a simple IDE disk is a lot more straightforward than
    emulating a SCSI-controller + disk. Now, I want to know whether this
    should work (which means that there's a bug I need to find), or
    whether I should use my time to implement SCSI emulation.

    So here's my question:

    Is a real ES40 able to boot OpenVMS from an IDE (ATA) disk? Has anyone
    ever tried this? Please let me know if you have.

    Thanks,

    Camiel Vanderhoeven

    http://www.camicom.com/es40
    http://www.sourceforge.net/projects/es40


  2. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?


    > Is a real ES40 able to boot OpenVMS from an IDE (ATA) disk? Has anyone
    > ever tried this? Please let me know if you have.


    Yes. We have booted an ES40 from an internal Maxtor 51024U2 on an IDE
    bus. IDE disks show up as DQxx devices (OpenVMS 7.3-1).

    ---------------------------------------------------------
    Tom Wade | EMail: tee dot wade at eurokom dot ie
    EuroKom | Tel: +353 (1) 296-9696
    A2, Nutgrove Office Park | Fax: +353 (1) 296-9697
    Rathfarnham | Disclaimer: This is not a disclaimer
    Dublin 14 | Tip: "Friends don't let friends do Unix !"
    Ireland

  3. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    In article <1172143525.297588.189480@h3g2000cwc.googlegroups.c om>, "Camiel" writes:
    >
    > jumping to bootstrap code
    > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910
    >


    %SYSTEM-W-NOSUCHFILE

    Does the OS version you're trying this with support the processor
    you're emulating?


  4. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    On Feb 22, 5:25 am, "Camiel" wrote:
    > Hi,
    >
    > I'm running into a problem with an emulator for the ES40 that I'm
    > working on. I get the following error:
    >
    > jumping to bootstrap code
    > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910


    Booting from an IDE disk should work fine. The error you are getting
    is a "file not found", so I would guess that the boot block does not
    properly identify the location of SYSBOOT.EXE. You will need to mount
    your disk on another OpenVMS system and run WRITEBOOT.EXE in order to
    update the boot block with the correct location of SYSBOOT.EXE.


  5. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?


    wrote in message
    news:1172160604.725647.133730@l53g2000cwa.googlegr oups.com...

    >> %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    >> %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    >
    > Booting from an IDE disk should work fine. The error you are getting
    > is a "file not found", so I would guess that the boot block does not
    > properly identify the location of SYSBOOT.EXE.


    Doesn't the boot block indicate the primary bootstrap, APB.EXE?
    It should find the SYSBOOT through the normal filesystem if the
    boot flags are correct.



  6. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    On Feb 22, 5:10 pm, dav...@montagar.com wrote:
    > On Feb 22, 5:25 am, "Camiel" wrote:
    >
    > > Hi,

    >
    > > I'm running into a problem with an emulator for the ES40 that I'm
    > > working on. I get the following error:

    >
    > > jumping to bootstrap code
    > > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    >
    > Booting from an IDE disk should work fine. The error you are getting
    > is a "file not found", so I would guess that the boot block does not
    > properly identify the location of SYSBOOT.EXE. You will need to mount
    > your disk on another OpenVMS system and run WRITEBOOT.EXE in order to
    > update the boot block with the correct location of SYSBOOT.EXE.


    Yes it does. I'm trying to run OpenVMS 8.3 (installation CD image on
    an emulated hard-disk) on an emulated ES40 (EV68).

    Camiel.


  7. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Oops... I saw that my last reply was a response to a different post
    than the one I replied to.

    I'm booting OpenVMS 8.3 (7.3 gives a similar message) on an ES40, so
    the hardware I'm trying to emulate should be compatible with the OS.

    The primary bootstrap, APB.EXE, is found correctly. My emulator runs
    the actual SRM console and PAL code, so I'm not emulating that. The
    complete boot sequence looks like this:

    AlphaServer ES40 Console V7.2-1, built on Jun 9 2006 at 15:36:48

    CPU 0 booting

    (boot dqa0.0.0.15.0 -flags 0,0)
    block 0 of dqa0.0.0.15.0 is a valid boot block
    reading 1226 blocks from dqa0.0.0.15.0
    bootstrap code read in
    base = 200000, image_start = 0, image_bytes = 99400(627712)
    initializing HWRPB at 2000
    initializing page table at 7f56000
    initializing machine state
    setting affinity to the primary CPU
    jumping to bootstrap code
    %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    halted CPU 0

    halt code = 5
    HALT instruction executed
    PC = 20004048
    warning -- HWRPB is invalid



    Camiel Vanderhoeven

    http://www.camicom.com/es40
    http://www.sourceforge.net/projects/es40


  8. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Camiel wrote:
    >
    > Oops... I saw that my last reply was a response to a different post
    > than the one I replied to.
    >
    > I'm booting OpenVMS 8.3 (7.3 gives a similar message) on an ES40, so
    > the hardware I'm trying to emulate should be compatible with the OS.
    >
    > The primary bootstrap, APB.EXE, is found correctly. My emulator runs
    > the actual SRM console and PAL code, so I'm not emulating that. The
    > complete boot sequence looks like this:
    >
    > AlphaServer ES40 Console V7.2-1, built on Jun 9 2006 at 15:36:48
    >
    > CPU 0 booting
    >
    > (boot dqa0.0.0.15.0 -flags 0,0)
    > block 0 of dqa0.0.0.15.0 is a valid boot block
    > reading 1226 blocks from dqa0.0.0.15.0
    > bootstrap code read in
    > base = 200000, image_start = 0, image_bytes = 99400(627712)
    > initializing HWRPB at 2000
    > initializing page table at 7f56000
    > initializing machine state
    > setting affinity to the primary CPU
    > jumping to bootstrap code
    > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910
    >
    > halted CPU 0
    >
    > halt code = 5
    > HALT instruction executed
    > PC = 20004048
    > warning -- HWRPB is invalid
    >
    > Camiel Vanderhoeven
    >
    > http://www.camicom.com/es40
    > http://www.sourceforge.net/projects/es40


    Seems to me that somehow the disk geometry being used may not be correct.

    What exactly do you mean when you say "CD image"? ...and how was it created?

    Are you sure the geometry of the emulated disk matches the original CD?

    Please provide more info as this is likely where the hiccup will be found.

    --
    David J Dachtera
    dba DJE Systems
    http://www.djesys.com/

    Unofficial OpenVMS Marketing Home Page
    http://www.djesys.com/vms/market/

    Unofficial Affordable OpenVMS Home Page:
    http://www.djesys.com/vms/soho/

    Unofficial OpenVMS-IA32 Home Page:
    http://www.djesys.com/vms/ia32/

    Unofficial OpenVMS Hobbyist Support Page:
    http://www.djesys.com/vms/support/

  9. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Boot with flags 0,30000 to get more info, and post it here.

    Jur.

    Camiel wrote:
    > Oops... I saw that my last reply was a response to a different post
    > than the one I replied to.
    >
    > I'm booting OpenVMS 8.3 (7.3 gives a similar message) on an ES40, so
    > the hardware I'm trying to emulate should be compatible with the OS.
    >
    > The primary bootstrap, APB.EXE, is found correctly. My emulator runs
    > the actual SRM console and PAL code, so I'm not emulating that. The
    > complete boot sequence looks like this:
    >
    > AlphaServer ES40 Console V7.2-1, built on Jun 9 2006 at 15:36:48
    >
    > CPU 0 booting
    >
    > (boot dqa0.0.0.15.0 -flags 0,0)
    > block 0 of dqa0.0.0.15.0 is a valid boot block
    > reading 1226 blocks from dqa0.0.0.15.0
    > bootstrap code read in
    > base = 200000, image_start = 0, image_bytes = 99400(627712)
    > initializing HWRPB at 2000
    > initializing page table at 7f56000
    > initializing machine state
    > setting affinity to the primary CPU
    > jumping to bootstrap code
    > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910
    >
    > halted CPU 0
    >
    > halt code = 5
    > HALT instruction executed
    > PC = 20004048
    > warning -- HWRPB is invalid
    >
    >
    >
    > Camiel Vanderhoeven
    >
    > http://www.camicom.com/es40
    > http://www.sourceforge.net/projects/es40
    >


  10. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Hi Jur,

    boot -fl 0,30000 leads to a crash:

    P00>>>b -fl 0,30000
    (boot dqa0.0.0.15.0 -flags 0,30000)
    block 0 of dqa0.0.0.15.0 is a valid boot block
    reading 1226 blocks from dqa0.0.0.15.0
    bootstrap code read in
    base = 200000, image_start = 0, image_bytes = 99400(627712)
    initializing HWRPB at 2000
    initializing page table at 7f56000
    initializing machine state
    setting affinity to the primary CPU
    jumping to bootstrap code

    ************************************************** ***********
    * Exception taken before exception handler has been loaded! *
    * Unable to take crash dump! *
    ************************************************** ***********

    Access control violation through vector 00000080

    R00: 00000000 00000000 R01: 00000000 00000000 R02: 00000000
    20001228
    R03: 00000080 200621B0 R04: 00000000 200E0010 R05: 00000000
    00000000
    R06: 00000000 20096208 R07: 00000000 00000001 R08: 00000000
    20099202
    R09: 00000000 00000005 R10: 00000000 0000FFFF R11: 00000000
    00000005
    R12: 00000000 2006A298 R13: 00000000 00000002 R14: 00000000
    00000000
    R15: 00000000 200DFE12 R16: 00000000 00000005 R17: 00000000
    200DFE12
    R18: 00000000 2006A2A0 R19: 00000000 00000010 R20: 00000000
    00000066
    R21: 00000000 00000002 R22: 00000000 200DFE16 R23: 00303030
    30332C30
    R24: 00000000 2005EA20 R25: 00000000 200DFE13 R26: 00000000
    20096208
    R27: 00000000 20065630 R28: 00000000 200DFE10 R29: 00000000
    200DFC80

    R30: 200DFC40: 00000000 200652F8 Saved R02
    200DFC48: 00000000 2006A284 Saved R03
    200DFC50: 00000000 20098FF0 Saved R04
    200DFC58: 00000000 00000001 Saved R05
    200DFC60: 00000000 2006A28C Saved R06
    200DFC68: 00000000 00000001 Saved R07
    200DFC70: 00000000 2005E470 Exception PC
    200DFC78: 00000000 00001F00 Exception PS


    halted CPU 0

    halt code = 5
    HALT instruction executed
    PC = 20001358
    warning -- HWRPB is invalid
    P00>>>


  11. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    On Feb 23, 1:24 am, David J Dachtera
    wrote:
    > Camiel wrote:
    >
    > > Oops... I saw that my last reply was a response to a different post
    > > than the one I replied to.

    >
    > > I'm booting OpenVMS 8.3 (7.3 gives a similar message) on an ES40, so
    > > the hardware I'm trying to emulate should be compatible with the OS.

    >
    > > The primary bootstrap, APB.EXE, is found correctly. My emulator runs
    > > the actual SRM console and PAL code, so I'm not emulating that. The
    > > complete boot sequence looks like this:

    >
    > > AlphaServer ES40 Console V7.2-1, built on Jun 9 2006 at 15:36:48

    >
    > > CPU 0 booting

    >
    > > (boot dqa0.0.0.15.0 -flags 0,0)
    > > block 0 of dqa0.0.0.15.0 is a valid boot block
    > > reading 1226 blocks from dqa0.0.0.15.0
    > > bootstrap code read in
    > > base = 200000, image_start = 0, image_bytes = 99400(627712)
    > > initializing HWRPB at 2000
    > > initializing page table at 7f56000
    > > initializing machine state
    > > setting affinity to the primary CPU
    > > jumping to bootstrap code
    > > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    >
    > > halted CPU 0

    >
    > > halt code = 5
    > > HALT instruction executed
    > > PC = 20004048
    > > warning -- HWRPB is invalid

    >
    > > Camiel Vanderhoeven

    >
    > >http://www.camicom.com/es40
    > >http://www.sourceforge.net/projects/es40

    >
    > Seems to me that somehow the disk geometry being used may not be correct.
    >
    > What exactly do you mean when you say "CD image"? ...and how was it created?
    >
    > Are you sure the geometry of the emulated disk matches the original CD?
    >
    > Please provide more info as this is likely where the hiccup will be found.


    I created the image as follows: I took the original OpenVMS 8.3
    installation CD-ROM, and used Nero to extract an ISO image from it.
    This should result in a flat file, that has all the data on the CD in
    it. SRM accesses the disk in LBA (logical block addressing) mode, so
    geometry (cylinders, heads, sectors) should not be an issue. In fact,
    this seems to be demonstrated already by the fact that SRM manages to
    load the primary bootstrap image (APB.EXE).

    It could be that APB.EXE does not use LBA addressing to find
    SYSBOOT.EXE though... That's something worth checking out. I'll add
    some more debugging to the IDE emulator, and see if that is the case.

    Thanks,

    Camiel.

    >
    > --
    > David J Dachtera
    > dba DJE Systemshttp://www.djesys.com/
    >
    > Unofficial OpenVMS Marketing Home Pagehttp://www.djesys.com/vms/market/
    >
    > Unofficial Affordable OpenVMS Home Page:http://www.djesys.com/vms/soho/
    >
    > Unofficial OpenVMS-IA32 Home Page:http://www.djesys.com/vms/ia32/
    >
    > Unofficial OpenVMS Hobbyist Support Page:http://www.djesys.com/vms/support/




  12. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?


    "Camiel" wrote in message
    news:1172227848.363049.45770@s48g2000cws.googlegro ups.com...
    > On Feb 23, 1:24 am, David J Dachtera
    > wrote:


    >
    > I created the image as follows: I took the original OpenVMS 8.3
    > installation CD-ROM, and used Nero to extract an ISO image from it.
    > This should result in a flat file, that has all the data on the CD in
    > it. SRM accesses the disk in LBA (logical block addressing) mode, so
    > geometry (cylinders, heads, sectors) should not be an issue. In fact,
    > this seems to be demonstrated already by the fact that SRM manages to
    > load the primary bootstrap image (APB.EXE).
    >
    > It could be that APB.EXE does not use LBA addressing to find
    > SYSBOOT.EXE though... That's something worth checking out. I'll add
    > some more debugging to the IDE emulator, and see if that is the case.
    >
    > Thanks,
    >


    APB is located by block address, SYSBOOT is located by filesystem lookup.

    Mount the disk on a VMS system and see if you have SYSBOOT on it and a valid
    file structure.




  13. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    On Feb 23, 2:44 pm, "FredK" wrote:
    > "Camiel" wrote in message
    >
    > news:1172227848.363049.45770@s48g2000cws.googlegro ups.com...
    >
    >
    >
    > > On Feb 23, 1:24 am, David J Dachtera
    > > wrote:

    >
    > > I created the image as follows: I took the original OpenVMS 8.3
    > > installation CD-ROM, and used Nero to extract an ISO image from it.
    > > This should result in a flat file, that has all the data on the CD in
    > > it. SRM accesses the disk in LBA (logical block addressing) mode, so
    > > geometry (cylinders, heads, sectors) should not be an issue. In fact,
    > > this seems to be demonstrated already by the fact that SRM manages to
    > > load the primary bootstrap image (APB.EXE).

    >
    > > It could be that APB.EXE does not use LBA addressing to find
    > > SYSBOOT.EXE though... That's something worth checking out. I'll add
    > > some more debugging to the IDE emulator, and see if that is the case.

    >
    > > Thanks,

    >
    > APB is located by block address, SYSBOOT is located by filesystem lookup.
    >
    > Mount the disk on a VMS system and see if you have SYSBOOT on it and a valid
    > file structure.


    Hi Fred,

    I can't mount the disk on a VMS system since it's a virtual disk (i.e.
    just a file on my Linux or Windows machine's filesystem)

    Camiel.


  14. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?


    "Camiel" wrote in message
    news:1172240504.926611.315790@p10g2000cwp.googlegr oups.com...
    > On Feb 23, 2:44 pm, "FredK" wrote:
    >> "Camiel" wrote in message
    >>
    >> news:1172227848.363049.45770@s48g2000cws.googlegro ups.com...
    >>
    >>
    >>
    >> > On Feb 23, 1:24 am, David J Dachtera
    >> > wrote:

    >>
    >> > I created the image as follows: I took the original OpenVMS 8.3
    >> > installation CD-ROM, and used Nero to extract an ISO image from it.
    >> > This should result in a flat file, that has all the data on the CD in
    >> > it. SRM accesses the disk in LBA (logical block addressing) mode, so
    >> > geometry (cylinders, heads, sectors) should not be an issue. In fact,
    >> > this seems to be demonstrated already by the fact that SRM manages to
    >> > load the primary bootstrap image (APB.EXE).

    >>
    >> > It could be that APB.EXE does not use LBA addressing to find
    >> > SYSBOOT.EXE though... That's something worth checking out. I'll add
    >> > some more debugging to the IDE emulator, and see if that is the case.

    >>
    >> > Thanks,

    >>
    >> APB is located by block address, SYSBOOT is located by filesystem lookup.
    >>
    >> Mount the disk on a VMS system and see if you have SYSBOOT on it and a
    >> valid
    >> file structure.

    >
    > Hi Fred,
    >
    > I can't mount the disk on a VMS system since it's a virtual disk (i.e.
    > just a file on my Linux or Windows machine's filesystem)
    >


    Then I have no suggestion for you, other than to guess that whatever you
    have on this "virtual disk" is corrupt as far as the filesystem goes. You
    need another way to create the virtual disk.




  15. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    On Feb 23, 11:50 am, "Camiel" wrote:
    > On Feb 23, 1:24 am, David J Dachtera
    > wrote:
    >
    >
    >
    > >Camielwrote:

    >
    > > > Oops... I saw that my last reply was a response to a different post
    > > > than the one I replied to.

    >
    > > > I'm booting OpenVMS 8.3 (7.3 gives a similar message) on an ES40, so
    > > > the hardware I'm trying to emulate should be compatible with the OS.

    >
    > > > The primary bootstrap, APB.EXE, is found correctly. My emulator runs
    > > > the actual SRM console and PAL code, so I'm not emulating that. The
    > > > complete boot sequence looks like this:

    >
    > > > AlphaServer ES40 Console V7.2-1, built on Jun 9 2006 at 15:36:48

    >
    > > > CPU 0 booting

    >
    > > > (boot dqa0.0.0.15.0 -flags 0,0)
    > > > block 0 of dqa0.0.0.15.0 is a valid boot block
    > > > reading 1226 blocks from dqa0.0.0.15.0
    > > > bootstrap code read in
    > > > base = 200000, image_start = 0, image_bytes = 99400(627712)
    > > > initializing HWRPB at 2000
    > > > initializing page table at 7f56000
    > > > initializing machine state
    > > > setting affinity to the primary CPU
    > > > jumping to bootstrap code
    > > > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE
    > > > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910

    >
    > > > halted CPU 0

    >
    > > > halt code = 5
    > > > HALT instruction executed
    > > > PC = 20004048
    > > > warning -- HWRPB is invalid

    >
    > > >CamielVanderhoeven

    >
    > > >http://www.camicom.com/es40
    > > >http://www.sourceforge.net/projects/es40

    >
    > > Seems to me that somehow the disk geometry being used may not be correct.

    >
    > > What exactly do you mean when you say "CD image"? ...and how was it created?

    >
    > > Are you sure the geometry of the emulated disk matches the original CD?

    >
    > > Please provide more info as this is likely where the hiccup will be found.

    >
    > I created the image as follows: I took the original OpenVMS 8.3
    > installation CD-ROM, and used Nero to extract an ISO image from it.
    > This should result in a flat file, that has all the data on the CD in
    > it. SRM accesses the disk in LBA (logical block addressing) mode, so
    > geometry (cylinders, heads, sectors) should not be an issue. In fact,
    > this seems to be demonstrated already by the fact that SRM manages to
    > load the primary bootstrap image (APB.EXE).
    >
    > It could be that APB.EXE does not use LBA addressing to find
    > SYSBOOT.EXE though... That's something worth checking out. I'll add
    > some more debugging to the IDE emulator, and see if that is the case.
    >


    I added some more debugging, and all disc-accesses while APB is
    running are done in LBA mode, one sector at a time, so that's not it.
    The information APB.EXE reads from the disk looks a lot like a VMS
    directory to me. (I see stuff like SYS0.DIR)

    Camiel.


  16. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Bob Koehler wrote:
    > In article <1172143525.297588.189480@h3g2000cwc.googlegroups.c om>, "Camiel" writes:
    >>
    >> jumping to bootstrap code
    >> %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE


    You need to boot that node as a satelline. And then use the WRITEBOOT
    utility to cause the locationn of the SYSBOOT.EXE file to be written to
    some specific sectors at the top of the disk drive so that the hardware can
    find and load the code.

  17. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    JF Mezei wrote:
    > Bob Koehler wrote:
    >> In article <1172143525.297588.189480@h3g2000cwc.googlegroups.c om>,
    >> "Camiel" writes:
    >>>
    >>> jumping to bootstrap code
    >>> %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE

    >
    > You need to boot that node as a satelline. And then use the WRITEBOOT
    > utility to cause the locationn of the SYSBOOT.EXE file to be written to
    > some specific sectors at the top of the disk drive so that the hardware
    > can find and load the code.


    Um, WRITEBOOT and SET BOOTBLOCK operate on references from LBN0 to
    (in the case of OpenVMS Alpha) APB.EXE, and not on SYSBOOT.EXE. APB.EXE
    is the primary bootstrap file for OpenVMS Alpha, and -- once loaded by
    the SRM console -- it then uses an in-built file system and various data
    structures and driver hooks to locate and load the secondary bootstrap
    image, SYSBOOT.EXE. This particular bootstrap sequence is already well
    past the point of the bootblock references and of loading APB.EXE, too,
    as APB.EXE is already looking around for the secondary bootfile. This
    is the primitive file system and the console and related bootstrap
    driver code that are in flight -- and failing -- here.

    On no evidence, my bet here is on Nero. Either the tool itself, or
    the particular command sequence involved. Replicate the disk on OpenVMS
    itself using COPY on a foreign-mounted CD, or by using CDDVD/COPY on
    V8.3 and later (SET COMMAND SYS$ETC:CDDVD$TOOLS first) or use a tool
    such as the Unix DD tool to replicate the disk.

    The "for grins" test here: burn a CD disk from this image, and see if
    a real OpenVMS box "likes" it, or if ANALYZE kicks out any errors, or if
    SYSBOOT is missing or otherwise unreachable.

    Various of the PC tools will happily not do what exactly you expect
    in this area, based on long experience chasing these cases. In one
    case, the PC tool was happily chunking off the first 128 blocks or so
    during the recording, as that was clearly not valid on any CD or so the
    authors of the tool had assumed. It actually ignored everything in
    front of where the ISO-9660 structures are located. The tough part here
    is that some versions and some version levels of these PC recording
    tools can work just fine, and others can be really weird. And there's
    no consistent terminology across the tools.

    If you use a Microsoft Windows box, one of the few tools I've found
    reasonably functional in this regard is the CDBurnerXP package.
    It's also free. And it's also pretty easy
    to use.

    The other issue here can be due to the particular emulation of the
    AlphaServer ES40 box, as that's a non-trivial undertaking. Hardware
    bugs can be rather cryptic, and emulators present hardware bugs. And
    without the APB bootstrap sequence and processing description from the
    IDSM and without the APB and related driver interface source listings,
    this case of emulation failure is even more interesting and difficult to
    troubleshoot and to resolve.

    --
    www.HoffmanLabs.com
    Services for OpenVMS

  18. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    Stephen Hoffman wrote:
    > Um, WRITEBOOT and SET BOOTBLOCK operate on references from LBN0 to
    > (in the case of OpenVMS Alpha) APB.EXE, and not on SYSBOOT.EXE.


    Thanks for the correction.

    > as APB.EXE is already looking around for the secondary bootfile. This
    > is the primitive file system and the console and related bootstrap
    > driver code that are in flight -- and failing -- here.


    If the initial reference to APB.EXE was correctly written to this disk,
    wouldn't there be some assumption that the rest of the disk would also be
    correctly loaded ?

    How was this particular device populated in the first place ?


    I remember my Vaxstation 3100 having problem with an old SCSI-1 disk. It
    would begin the boot, and then just hang because it was waiting for a SCSI
    response from a disk that didn't understand the question. Physically
    unplugging the power from drive and plugging it back in would clear the jam
    and allow the boot to continue. (and no problems after that).

    Is it possible that the "firmware" is able to access the device and begin
    the boot sequence, but that once the VMS drivers take over, it causes
    problems ? (remember that his IDE drive is emulated so perhaps VMS makes
    some slightly different IDE requests that his IDE emulator doesn't quite
    handle properly ?

    Note to the original poster: DS10 class machines have no problems booting
    from IDE drives. This is pretty standard.

  19. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    JF Mezei wrote:
    > Stephen Hoffman wrote:
    >> Um, WRITEBOOT and SET BOOTBLOCK operate on references from LBN0 to
    >> (in the case of OpenVMS Alpha) APB.EXE, and not on SYSBOOT.EXE.

    >
    > Thanks for the correction.
    >
    >> as APB.EXE is already looking around for the secondary bootfile. This
    >> is the primitive file system and the console and related bootstrap
    >> driver code that are in flight -- and failing -- here.

    >
    > If the initial reference to APB.EXE was correctly written to this disk,
    > wouldn't there be some assumption that the rest of the disk would also
    > be correctly loaded ?


    You can assume whatever you like, but didn't some local systems get
    scrozzled by some nasty evil corruptions not to long ago.

    > I remember my Vaxstation 3100 having problem with an old SCSI-1 disk. It
    > would begin the boot, and then just hang because it was waiting for a
    > SCSI response from a disk that didn't understand the question.
    > Physically unplugging the power from drive and plugging it back in would
    > clear the jam and allow the boot to continue. (and no problems after that).


    There are days I don't understand the question. Usually more coffee
    helps. But I digress. :-)

    There is a non-trivial difference from device to device, whether we
    are discussing SCSI controllers, SCSI disks, IDE bus interfaces, or ATA
    or ATAPI IDE targets. CD and DVD devices are particularly familiar, and
    I've yet to meet two that are identical. Some vendors are better than
    others here, but device-level compatibility is far from assured.

    Firmware-based devices such as SCSI disks can quite commonly head off
    into their own little world. But it's OK, the SCSI deities know them
    there. :-) Drivers can get confused. Firmware can get confused. The
    VAXstation 3100 series firmware some of the earliest-generation SCSI
    support in the product line, too.

    The old saw certainly applies: all usage of the term "compatible" to
    describe a widget should be interpreted as meaning "different", as if
    the widgets involved were truly identical, they would be referred to as
    being "identical".


    > Is it possible that the "firmware" is able to access the device and
    > begin the boot sequence, but that once the VMS drivers take over, it
    > causes problems ? (remember that his IDE drive is emulated so perhaps
    > VMS makes some slightly different IDE requests that his IDE emulator
    > doesn't quite handle properly ?


    The whole box is emulated.

    That is orders of magnitude larger than an emulated IDE ATA disk
    interface. This could be something in the structures leading back to
    the controller that's not quite right, or an instruction-level error, or
    most anything else that could conceivably be wrong with "hardware".

    APB has not gotten as far along as the "real" OpenVMS drivers here.
    APB and SYSBOOT both use APB-based drivers. Once SYSBOOT is rolling,
    then more advanced drivers are loaded and activated.


    --
    www.HoffmanLabs.com
    Services for OpenVMS

  20. Re: Is it possible to boot OpenVMS from an IDE disk on an ES40?

    In article <1172240504.926611.315790@p10g2000cwp.googlegroups. com>, "Camiel" writes:
    >I can't mount the disk on a VMS system since it's a virtual disk (i.e.
    >just a file on my Linux or Windows machine's filesystem)


    No excuse, since VMS has the logical disk driver (since about a decade
    as freeware, and now as part of OpenVMS V8)

    $ @SYS$STARTUP:LD$STARTUP ! often done in VMS startup - see LDA0 template
    $ LD CONN vms083-ods2-image.dsk LDA1 ! after transferring the file to VMS
    $ MOUNT LDA1 ALPHA083
    $ ANAL/DISK/REPA LDA1: ! /REPAIR only if needed
    $ MC WRITEBOOT
    Update VAX portion of boot block (default is Y) : N
    Update AXP portion of boot block (default is Y) : Y
    Enter AXP boot file : LDA1:[VMS$COMMON.SYSEXE]APB.EXE

    YMMV

    --
    Peter "EPLAN" LANGSTOEGER
    Network and OpenVMS system specialist
    E-mail peter@langstoeger.at
    A-1030 VIENNA AUSTRIA I'm not a pessimist, I'm a realist

+ Reply to Thread
Page 1 of 3 1 2 3 LastLast