How to erase a faulty flashbank ? - Hewlett Packard

This is a discussion on How to erase a faulty flashbank ? - Hewlett Packard ; Hello John, the source of the PVARS command was posted in this newsgroup by Cyrille. However it is MASD syntax which I never got familiar with and I never got it to run for Port 2. But from memory (IIRC) ...

+ Reply to Thread
Page 2 of 4 FirstFirst 1 2 3 4 LastLast
Results 21 to 40 of 66

Thread: How to erase a faulty flashbank ?

  1. xSTO does not handle storing in Port 2 properly

    Hello John,

    the source of the PVARS command was posted in this newsgroup by
    Cyrille.

    However it is MASD syntax which I never got familiar with and I never
    got it to run for Port 2. But from memory (IIRC) free memory is
    reported what *could* be stored in flash. So "free" space is
    calculated as the sum of all flash memory minus the objects stored in
    flash which are not marked as deleted and have a valid crc.
    Hopefully this calculation is less buggy then the attempt to store in
    flash.

    Annother interesting thing I am just noticing is the following:
    Assume that all flashbanks are filled up with objects that are marked
    as beeing deleted.
    Now through a sys-rpl program try to store a large object ( >100 kb)
    in port 2.
    It will work (assuming enough ram) but even if the xSTO is surrounded
    by ERRSET/ERRTRAP the calc warmstarts after storing the file.
    Maybe someone can confirm this behavior, I am running ROM #2.09.

    I guess I have to add something to the "infamous" STO-BUG.

    What is the comparison of infamous ?

    Greetings
    Andreas
    http://www.software49g.gmxhome.de


    On 9 Nov., 00:14, "John H Meyers" wrote:
    > On Thu, 08 Nov 2007 13:45:56 -0600, Claudio Lapilli wrote:
    > > Find the bank with smallest free space that can contain the lib.

    >
    > Is "free" space calculated in terms of:
    >
    > o Total space not occupied by non-deleted objects?
    >
    > o Remaining space between "current end of data" and "end of bank"?
    >
    > Consider a bank almost filled with deleted objects, for example;
    > the first measure of "free" space in that bank would be high,
    > while the second measure would be low.
    >
    > Thanks.




  2. Re: xSTO does not handle storing in Port 2 properly

    On Thu, 08 Nov 2007 17:33:54 -0600:

    Thanks for your reply, Andreas Möller.

    I may not have clarified, by quoting Claudio's prior post,
    that I'm trying to find out what "free space" means,
    only in the context of the logic of the STO command,
    as just explained by Claudio, where the term "free space"
    might sensibly have either meaning, but I would guess
    that the simpler (second) logic is better; i.e. look first
    for a flash bank which doesn't need re-organizing,
    so as not to wastefully keep erasing banks
    when a STO might be completed in another bank
    without performing an erase cycle, nor going through
    a re-organization, which takes much longer,
    and can fail if free memory is insufficient.

    The original description simply was slightly fuzzy
    about defining "free space,"
    which may not be in terms of what PVARS computes,
    but rather the lesser amount of space left
    before an erase cycle is needed.

    The best algorithms for distributing items in flash
    might differ; say, did Knuth ever work on this problem?
    (or is it really simple enough for mere mortals to solve?

    [r->] [OFF]

  3. Re: Does xSTO handles storing in Port 2 properly ?

    Hello Claudio,

    > It's called ARMCODE object, and it already exists. It can be

    EVAL'ed
    > just like a regular CODE object


    Please do not forget a size field behind the prolog for easy and fast
    skipping ;-)

    Greetings
    Andreas




  4. Re: ARMCODE object

    On Thu, 08 Nov 2007 17:27:23 -0600, Claudio Lapilli wrote:

    > ARMCODE object


    Has it a new prolog, or is it embedded within a CODE object?
    (the latter would avoid potential problems brought up below)

    If new prolog, can a Saturn-based calc (HP49G or Emu48/Debug4x)
    evaluate it with a graceful error (or NOP), rather than crashing?

    If embedded within a list (or program) that is in turn
    transferred to a "Saturn-based" calculator,
    can SKIPOB skip it, also without crashing?

    What type will CK&DISPATCHx etc. consider it to be?

    [r->] [OFF]

  5. Re: ARMCODE object

    Oh, some parts were previously answered:

    > It does run truly seamlessly in sysRPL.
    > It does not run within a Saturn CODE object,
    > since it's a different type of object.


    > It can be embedded in a sysRPL program with no problems.


    "embedded in a sysRPL program" only when using a new ROM?

    The questions raised about objects inadvertently (or "adverdently"
    transferred to other calculators still remain,
    perhaps with different answers for 49G+/50G/48Gii (with 2.09, say)
    than for "pure Saturn" 49G or emulators?

    > In any case, you can do in ARM
    > almost anything you can do in Saturn assembler,
    > so why would you need to embed ARM in Saturn?


    Because existing Saturn programs can already do one or two things
    quite satisfactorily, based on over 15 years of development,
    so why "re-write the wheel" for everything already developed?

    Crunch the big matrix in ARMCODE (in fact, create a few "standard"
    function replacements, such as matrix inversion, SolveSys, SORT, etc.)
    but do symbolics and interface with the display, the user,
    variables, alarms, key definitions, and everything else in SysRPL
    (or even UserRPL, heaven forbid

    Or perhaps someone is re-writing all of UserRPL/SysRPL/CAS
    in ARMCODE? Then a previously asked question is answered,
    and the entire ROM has been rewritten in "pure ARM"; my,
    that was fast work! (Did you retain the Supported entry points?

    [r->] [OFF]

  6. Re: ARMCODE object

    On Sat, 10 Nov 2007 07:14:10 -0600:

    > What type will CK&DISPATCHx etc. consider it to be?


    Was there any unused pre-existing type?

    If so, was it already friendly to SKIPOB and EVAL?

    Thanks for fielding all the curious questions,
    and best wishes.

    [r->] [OFF]

  7. Re: Does xSTO handles storing in Port 2 properly ?

    On Nov 8, 6:14 pm, "John H Meyers" wrote:
    > On Thu, 08 Nov 2007 13:45:56 -0600, Claudio Lapilli wrote:
    > > Find the bank with smallest free space that can contain the lib.

    >
    > Is "free" space calculated in terms of:
    >
    > o Total space not occupied by non-deleted objects?
    >
    > o Remaining space between "current end of data" and "end of bank"?
    >
    > Consider a bank almost filled with deleted objects, for example;
    > the first measure of "free" space in that bank would be high,
    > while the second measure would be low.
    >
    > Thanks.


    When trying to STO an object, free space is between end-of-data and
    end-of-bank.
    When reported by the filer, free space = size of bank minus the sum of
    all non-deleted objects.

    Claudio


  8. Re: Does xSTO handles storing in Port 2 properly ?

    On Nov 8, 8:43 pm, "manjo" wrote:
    <...>
    >
    > Very nice,
    > Good work !!! :-)
    > so you're actualy sugesting we would use flash for code and RAM as
    > variables (data) segment and temporary memory allocation/use.
    > -i belive i can live with that


    The idea is to be very flexiible. If your code (ARMCODE object) is in
    RAM, then a secondary copy is created, properly aligned, and outside
    the Saturn RAM space. This is regardless of your code being previously
    aligned or not.
    Why make a copy if your code was already aligned? Simple: hpgcc3 will
    include more advanced saturn support, so you can STO in Home, create
    objects in TEMPOB, and even do garbage collection all from within your
    C program. This means that your ARMCODE object is very likely to be
    'moved around', just as it is in the Saturn world. The RPL environment
    takes care of moving objects, but in ARM we don't have that luxury. So
    it's best to get it out of the Saturn way.
    If your code is in flash AND embedded in a library object (as a
    separate XLIB), then 2 things may happen: 1) If it's properly aligned,
    it's executed in-place. or 2) Not aligned, then a copy is created in
    RAM.

    You never know from where your code is executed, and you don't really
    need to know.


    >
    > -RAM as data segment and memory (heap) during execution


    RAM can be data and/or code segment

    > -Flash for code


    Flash is obviously for code, but also for read-only data (strings and
    so).

    <...>

    >
    > I'm not sure how memory allocation is done in Saturn environment.
    > So i figured there would be maximum 8 bytes of slack per ARMCODE object
    > which, i think, is reasonable price for ability to get it runing instantly.


    But you are thinking assembler-only. For a fully implemented
    executable, you need a separate address space, so you don't have to
    write relocatable code. This can only be achieved in page-aligned
    addresses, so it's 1kbyte or 4kbytes.
    And now that you mentioned 'instantly', a C program running in a tight
    loop executes one iteration every 7 ms, including the UserRPL
    overhead.
    (tested with <<1. 1000. FOR I MyProg NEXT >>)

    I would say that's "instantly enough" for execution.

    <...>

    > Also Saturn and sysRPL opcodes take considerably less space
    > so i see the opportunity to preserve space when writing programs
    > with low demands on speed:


    We'll change your idea of "considerably less". But you'll have to see
    for yourself.

    Claudio


  9. Re: Does xSTO handles storing in Port 2 properly ?

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    manjo wrote:

    <..>

    > Also Saturn and sysRPL opcodes take considerably less space
    > so i see the opportunity to preserve space when writing programs
    > with low demands on speed:


    Hmm ... ???
    Why go for low speed for a "considerable" small sized program?
    One of our HPGCC3 revisited decnumber demos demonstrates, that high
    performance and tiny program size don't contradict.
    'FACT.EX3' calculates 5000! with 16326 significant digits in ~1.8 sec.
    The executable's size is exactly 472 bytes.

    Something better, anyone ?



    - --
    Ingo Blank

    http://hpgcc.org
    http://blog.hpgcc.org

    To reply by email, please substitute "spam" with my first name in the
    reply address.

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (MingW32)
    Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

    iD8DBQFHNgmar9bi0BQTf30RAiyXAKCim5rf/Q+fxf28hEKVbzjcm7jSZACfdXUC
    FHM55UWQgLV+JHjSGUdEVdE=
    =iO7L
    -----END PGP SIGNATURE-----

  10. Re: Does xSTO handles storing in Port 2 properly ?

    > include more advanced saturn support, so you can STO in Home, create

    When you say store in HOME, do you mean in just the root or can you
    store inside subdirectories? I am most interested in storing inside
    the hidden directory as that is where I tend to put all my stuff.

    TW



  11. Re: Does xSTO handles storing in Port 2 properly ?

    On Sat, 10 Nov 2007 07:52:15 -0600, Claudio Lapilli wrote:

    > When trying to STO an object [in flash],
    > "free space" is between end-of-data and end-of-bank.


    Thought so. Thanks!

    > When reported by the filer, free space is
    > size of bank minus the sum of all non-deleted objects.


    What with the U.S. Dollar falling,
    "size of bank" is decreasing all the time, around here

    [r->] [OFF]

  12. Re: ARMCODE object

    On Nov 10, 8:14 am, "John H Meyers" wrote:
    > On Thu, 08 Nov 2007 17:27:23 -0600, Claudio Lapilli wrote:
    > > ARMCODE object

    >
    > Has it a new prolog, or is it embedded within a CODE object?
    > (the latter would avoid potential problems brought up below)


    1) It's not embedded. It's a separate object type.
    2) It's not exactly a new object, but an implementation of an old do-
    nothing object in rom (most DOEXTn are still unused)


    >
    > If new prolog, can a Saturn-based calc (HP49G or Emu48/Debug4x)
    > evaluate it with a graceful error (or NOP), rather than crashing?


    It will push the contents on the stack, just like a regular non-
    executable object. No error, no crash (but not exactly a NOP).
    A sysRPL program may expect a result on the stack, or may not expect
    the program to be pushed on the stack, so it may cause a crash as a
    collateral effect, but not directly related to the evaluation of the
    object.

    >
    > If embedded within a list (or program) that is in turn
    > transferred to a "Saturn-based" calculator,
    > can SKIPOB skip it, also without crashing?


    No issues with SKIPOB, since this object exists since the 48 series. I
    just renamed it and changed the EVAL behavior.
    The object can be embedded in lists and sysRPL programs, and any other
    composite object without problems.

    >
    > What type will CK&DISPATCHx etc. consider it to be?


    DOEXT4 prolog, shown as EXT3 in MASD.
    I gave it the name ARMCODE, and it shows the string "ARM/C Code" when
    pushed on the stack (it will show "External" on a machine that doesn't
    have hpgcc3 installed).

    Claudio


  13. Re: ARMCODE object

    On Nov 10, 8:34 am, "John H Meyers" wrote:
    <...>
    >
    > > In any case, you can do in ARM
    > > almost anything you can do in Saturn assembler,
    > > so why would you need to embed ARM in Saturn?

    >
    > Because existing Saturn programs can already do one or two things
    > quite satisfactorily, based on over 15 years of development,
    > so why "re-write the wheel" for everything already developed?
    >
    > Crunch the big matrix in ARMCODE (in fact, create a few "standard"
    > function replacements, such as matrix inversion, SolveSys, SORT, etc.)
    > but do symbolics and interface with the display, the user,
    > variables, alarms, key definitions, and everything else in SysRPL
    > (or even UserRPL, heaven forbid


    But manjo is talking about Saturn assembler. You don't get to use the
    CAS from assembler (mostly in FPTR's, so you need sysRPL). Almost
    anything you can do in Saturn assembler can be done in ARM assembler.
    Even better, it can be done in C.
    You don't use your alarms from assembler, nor the key definitions.
    There are only very minor things that you could need to use Saturn
    assembler for, and for very specialized (rare) apps.

    >
    > Or perhaps someone is re-writing all of UserRPL/SysRPL/CAS
    > in ARMCODE?


    Not that I know.

    Claudio


  14. Re: Does xSTO handles storing in Port 2 properly ?

    On Nov 10, 4:35 pm, TW wrote:
    > > include more advanced saturn support, so you can STO in Home, create

    >
    > When you say store in HOME, do you mean in just the root or can you
    > store inside subdirectories? I am most interested in storing inside
    > the hidden directory as that is where I tend to put all my stuff.
    >
    > TW


    I mean HOME as a generic place (main calculator ram). Of course you
    are able to STO/RCL in any subdirectory, including the hidden one.

    Claudio


  15. Re: Does xSTO handles storing in Port 2 properly ?

    > Claudio

    Just a thought I've been wondering about. I assume HP legal would
    have an issue with you redistributing a modified ROM image. Is the
    plan just to have a little "patcher" utility that will direct you to
    the HP site to download the ROM, then let you select a ROM file and
    then stick in the changes?

    TW


  16. Re: Does xSTO handles storing in Port 2 properly ?

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    TW wrote:
    >> Claudio

    >
    > Just a thought I've been wondering about. I assume HP legal would
    > have an issue with you redistributing a modified ROM image.


    I believe your own legal issues with selling GPL protected software,
    which is using the HPGCC libraries, are reason enough to worry.

    There are no obvious reasons by the way, why we would consider to
    discuss legal or other things regarding HPGCC3 with you or in a public
    forum.


    - --
    Ingo Blank

    http://hpgcc.org
    http://blog.hpgcc.org

    To reply by email, please substitute "spam" with my first name in the
    reply address.

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (MingW32)
    Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

    iD8DBQFHN6ihr9bi0BQTf30RArIeAJ9y8Amw0DXgXnzahk0hCL SW4/SUSwCgwu2Z
    9n5vId88MHefMH1hFjNm/yY=
    =XXyJ
    -----END PGP SIGNATURE-----

  17. Re: Does xSTO handles storing in Port 2 properly ?

    Ingo Blank wrote:
    >
    > I believe your own legal issues with selling GPL protected software,
    > which is using the HPGCC libraries, are reason enough to worry.


    I'm confused here - why would there be any problem selling GPL licensed
    software that uses GPL licensed libraries (HPGCC)? None of this violates
    GPL.

    - Kurt

  18. Re: Does xSTO handles storing in Port 2 properly ?

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    ~kurt wrote:
    > Ingo Blank wrote:
    >> I believe your own legal issues with selling GPL protected software,
    >> which is using the HPGCC libraries, are reason enough to worry.

    >
    > I'm confused here - why would there be any problem selling GPL licensed
    > software that uses GPL licensed libraries (HPGCC)? None of this violates
    > GPL.
    >


    Not the *selling* per se is the problem, at least not in terms of the
    GPL. I think major problems arise from civil law, when you sell a
    product, which you reject any warranties for, which does the GPL.

    But selling it as *closed source* without making available the complete
    sources violates the GPL, that's the point.

    Reason enough to worry, like I said before.

    The HPGCC license states clearly:

    This entire project, including the libraries, is released under the GPL.
    Under the terms of the GPL, you are required to release any software
    that uses these libraries under the GPL as well.

    Understandably, a large number of developers would prefer to keep their
    source code private. To solve this issue, additional rights are granted
    to non-profit developers. For non profit use only, you may use HPGCC
    without putting your code under the GPL


    And seriously, how would you SELL a program, if everyone can download
    the sources and compile it?

    Do you pay for your Linux distros?
    [ just in case you are not a Windoze bigot :-) ]


    - --
    Ingo Blank

    http://hpgcc.org
    http://blog.hpgcc.org

    To reply by email, please substitute "spam" with my first name in the
    reply address.

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (MingW32)
    Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

    iD8DBQFHOEvvr9bi0BQTf30RAvBwAJ0ak/pcwe7jPs3na12MxJA5C53sMQCaApxU
    ioIoiQ6JkoU3wES60UmC4dw=
    =dwrW
    -----END PGP SIGNATURE-----

  19. Re: Does xSTO handles storing in Port 2 properly ?

    > Not the *selling* per se is the problem, at least not in terms of the
    > GPL. I think major problems arise from civil law, when you sell a
    > product, which you reject any warranties for, which does the GPL.


    Due to the nature of our market, it wouldn't really matter if our
    software gave completely wrong answers. Being licensed professionals,
    surveyors themselves are responsible for the final product. If they
    use numbers spit out of a program blindly, they are liable - not the
    software developer. That is well established. Blaming someone else
    doesn't fly.

    Should things be spitting out wrong answers though, everyone would
    really quickly abandon the software and let everyone know about it.

    > But selling it as *closed source* without making available the complete
    > sources violates the GPL, that's the point.


    They are there for download with the updates, and they are on a hard
    medium of every unit that ships. What is in violation?

    > And seriously, how would you SELL a program, if everyone can download
    > the sources and compile it?


    Surveyors don't generally have the knowledge or desire to do so. They
    also want someone to be fixing problems and adding new instruments and
    new features.

    > Do you pay for your Linux distros?


    Me personally no. But a small local company I set up a server and
    several workstations all paid for RHEL at my encouragement. I run
    ubuntu on my laptop (used to be slack), XP on the desktop for the
    occasional game, and the wife has a macbook.

    However since you so succinctly stated last night, this isn't the
    place to discuss legal issues.

    TW




  20. Re: Does xSTO handles storing in Port 2 properly ?

    Since this threed is going off-topic
    i thought to give it a try and help make it entirely off topic.

    The work and effort of HPGCC team is *great*
    I admire your work Ingo, but be aware you're crossing the line here.

    First of if you plan to use a term *bigot* with windows, even if you did use
    ":-)"
    How to call the Linux bounch then ?
    Linux is like a church (religion) in its early stages.
    In these relations GPL is like a witch hunt, right ?
    It's like every Linux user signs a contract to convert several windows
    (obscure unworthy followers) to the right path of Linux in order to
    reach the Gates(maybe Bill) of Heaven or what ?

    I respect people willing to give away their work for free,
    but i don't belive in denying the right of people to get payed for their
    work,
    please don't try to make others to be same as you ...this will simply not be
    !!!

    while we're at linux vs. windows (again)
    (i will never get tired of this :-)

    as an example: my computer is brand name pc with windows (in price) so no
    special/external cost there (linux be aware)
    I am using the same machine for past 4 years with original installation
    (first installation).

    Even when i "break" something i always fixed it.
    The thing is that windows is so easy and fast to install that people don't
    bother fixing it.
    Reinstall seams faster and requires less care.
    The curse of windows was that careless user could mess with everything
    easily thus so many problems.

    Actualy i can "break" and fix my installation of windows (or choose to
    re-install)
    and i will loose much less time then just get Linux work on my machine.
    Not to mention that i really don't whish to be worried every single time
    when i buy a nice gadget
    if i will be able to make it work with Linux or not.

    Please don't try to "fix or update my opinion" windows update does the
    updating and fixing for
    me. I'm very pleased how it works, either make a little effort and keep
    installation updated myself or
    just make a check in the right checkbox and enjoy :-)
    People who run windows are *not* ignorant as most Linux users are trying to
    establish.
    We all have different motives for using computers thus different choice of
    OS.

    I'm not trying to pormote windows here, rather preach the respect for
    eachothers
    choices as i respect the choice of people runing Linux machines.

    To share or give away my work for free is also a choiche which needs care
    and understanding rather than roules, rape and violation !

    What does all this mean?
    -notthing, i guess it's just a glitch on my positronic Linux partition of
    the brain :-)

    --
    manjo
    http://fly.srk.fer.hr/~manjo/openfire
    | 49G+ | ROM 2.09 | hw serial:CN40213309 | sw serial:CN40701165 |



+ Reply to Thread
Page 2 of 4 FirstFirst 1 2 3 4 LastLast