GitLab

Detail:
Admin:
     jb ..

Version 1.47. Tagged as 'HdrSrc-1_47'

Version 1.46. Tagged as 'HdrSrc-1_46'

Version 1.45. Tagged as 'HdrSrc-1_45'

Detail:
  Merge of the latest 'Register' header files from Alan Glover.
Admin:
  The previous headers cotained faults, so were never checked into cvs.
  These ones are OK so I'm merging them.

Version 1.44. Tagged as 'HdrSrc-1_44'

* Changed Hdr:Debug not to rely on Hdr:Proc (to prevent some macro clashes).
* Added Hdr:Machine.STB5

Version 1.43. Tagged as 'HdrSrc-1_43'

Version 1.42. Tagged as 'HdrSrc-1_42'

Version 1.41. Tagged as 'HdrSrc-1_41'

Add Service_SCSIDriver

Version 1.40. Tagged as 'HdrSrc-1_40'

Version 1.39. Tagged as 'HdrSrc-1_39'

Version 1.38. Tagged as 'HdrSrc-1_38'

Detail:
  * Extra GraphicsV definitions
  * Service_Device renamed Service_Hardware
  * GRAB macro fixed for objasm compatibility
  * Enabled HiResTTX for Tungsten, and added 256-colour TTX option
Admin:
  Used in builds including release 5.00 (except the TTX bits).

Version 1.37. Tagged as 'HdrSrc-1_37'

Version 1.36. Tagged as 'HdrSrc-1_36'

Addition of Tungsten machine and Iyonix user interface.
New 1024k ROM image size.

Version 1.35. Tagged as 'HdrSrc-1_35'

  Changed some base 4 constants to binary (for Hdr2H)
Detail:
  Lots of additions to allocations since it's been a long time since
    the last resync with Pineapple.
Admin:
  E-mail received 2002/04/02 (includes hdr/RISCOS change in Kernel)


Version 1.34. Tagged as 'HdrSrc-1_34'

Detail:
  See above.
Admin:
  Requires Library-0_94. Required by DVBSubtitles.

Version 1.33. Tagged as 'HdrSrc-1_33'

Detail:
  This synchronisation includes some Pace allocations that have
    been made over the past few months, and includes the (textual comment
    only) notes on SWI OS allocations (ie. bits 20-23 of the SWI number)
Admin:
  E-mail received 7 January 2002.


Version 1.32. Tagged as 'HdrSrc-1_32'

Detail:
0xd1
Admin:
New version of PowerOn module works as expected.

Version 1.31. Tagged as 'HdrSrc-1_31'

Korean territory added.


Version 1.30. Tagged as 'HdrSrc-1_30'

Detail:
  DeviceFS has been enhanced with the ability to cooperate with Taskwindows
so as to allow multitasking to continue.  Any devicefs stream, if opened
with the special field entry 'sleep' will issue OS_UpCall 6 when it would
normally block.  A flag 'timeout' has also been added which will cause the
stream to issue a timeout error if it has slept for too long.

  These actions can be set using the generic IOCTLs 4 and 5, the former just
being a logical switch, the latter being the value in centiseconds for which
to sleep.

  There is also now a generic special field to specify non-blocking which may
be used as an alternative to the IOCTL.

Admin:
  Tested on a RiscPC.


Version 1.29. Tagged as 'HdrSrc-1_29'

Detail:
  This synchronisation includes some Pace allocations that have
    been made over the past few months.
Admin:
  E-mail received 8 August 2001.


Version 1.28. Tagged as 'HdrSrc-1_28'

Detail:

  * Ensured that the Lazarus standby button also applies to builds that
    use Machine=32.

Admin:

  * Untested.  Required for Customer T demonstrator.

Version 1.27. Tagged as 'HdrSrc-1_27'

Detail:
    A new file, 7168K, has been added to hdr.ImageSize.
Admin:
    Untested. Woohoo

Version 1.26. Tagged as 'HdrSrc-1_26'

Detail:
  Service_PRISM removed.
  Service_PRISM_Core replaces it as service &D3.
  Service_PRISM_Client added as service &D5.
Admin:
  Requested by Andrew to fix problem until new headers are received.


Version 1.25. Tagged as 'HdrSrc-1_25'

Admin:
  Details passed to Pineapple.


Version 1.24. Tagged as 'HdrSrc-1_24'

Detail:
  * Added register definitions for the new registers in those wacky new ARM2
    chips, r8_fiq and r9_fiq.
  * Added abort and undefined mode register definitions using the ARM ARM
    standardised suffix of _abt and _und (previously we only had _abort and
    _undef versions).
  * Added some sanity checking of the fields parameter to the RestPSR macro,
    to help warn the unwary of the effect of compiling No32bitCode.
Admin:
  Changes appear to function as desired. Debug builds of MPEGDriver actually
  require those extra register definitions.

Version 1.23. Tagged as 'HdrSrc-1_23'

Detail:

  * Martha uses a new user interface.  Provision
    made for Martha interface; type is set to "Embedded" since this
    is not a desktop system.

Admin:

  * None.

Version 1.22. Tagged as 'HdrSrc-1_22'

Detail:
    As above

Admin:
    Untested. Needed for Customer T builds.

Version 1.21. Tagged as 'HdrSrc-1_21'

Detail:
  A missing comma, and a short name that's too long, break builds.
  Fixed.
Admin:
  Built OK now - Pineapple mailed.

Version 1.20. Tagged as 'HdrSrc-1_20'

Detail:
  This synchronisation includes some Pace allocations that have
    been made over the past few months.
Admin:
  E-mail received 19 March 2001.

Version 1.19. Tagged as 'HdrSrc-1_19'

Detail:
  The PRMs also have this the wrong way around.
Admin:
  Updated file passed by to Pineapple to keep the master headers in step.

Version 1.18. Tagged as 'HdrSrc-1_18'

* New header file "Hdr:UserIF.UserIF" provided for definitions specific to the user interface on products.  This abstracts the UI definitions from the machine-specific ones, and (ultimately) allows more flexibility in the choice of Machine and UserIF build combinations.

Detail:

  * As part of the work to remove the (almost obsolete) STB flag, a new
    flag has been defined: Embedded_UI.  This is provided by the "UserIF"
    header file.  The file should not be directly included; instead use the
    same mechanism as for "Hdr:Machine.Machine" and use "Hdr:UserIF.<UserIF>".

    In the past, the STB flag has been used to mean different things in
    different software components, some machine-dependent but more often than
    not to differentiate desktop builds from STB/NC versions.  Two types
    of system are identified: "Desktop" and "Embedded"; UserIF_Type is
    defined as a string containing the UI type.  STB/NC user interface
    differences should be encapsulated in appropriate "Embedded_UI" switches.
    Any other user interface-specific changes should probably define a new
    flag in the UserIF header file.

    Presently, the STB flag remains intact in "Hdr:Machine.Machine" to allow
    builds to progress as normal until all software components that utilise
    it have been appropriately modified to remove the dependency.  For
    example, the Kernel makes heavy use of the flag, but after HAL-isation
    will no longer do so.  Once all software components have been modified
    the flag will be withdrawn, and feature tests or appropriately defined
    user interface flags will be used instead.

  * "Hdr:CMOS" revised to remove definitions under the control of the defunct
    ObsoleteNC1CMOS flag.  Furthermore, the STB flag conditional has been
    removed.

Admin:

  * Tested using appropriately modified software components on a 32-bit
    Lazarus build.

  * Use of the STB flag is now deprecated.  Software components should
    use "Hdr:UserIF.<UserIF>" or specific feature tests from this point
    onwards.  Appropriate header files exist inside the "Hdr.UserIF"
    directory for each user interface type.

Version 1.17. Tagged as 'HdrSrc-1_17'

Version 1.16. Tagged as 'HdrSrc-1_16'

Detail:
  OSLib defines "True" and "False" as integer variables, but Hdr:Symbols
  defines them as logical ones. Hdr:Proc assigned "True" and "False" to
  logical variables, so there was no way that it could be used with any
  component that used any OSLib or even DefMod-generated headers. Hdr:Proc
  now uses the unambiguous "{TRUE}" and "{FALSE}" symbols.
Admin:
  Required by SoundFileExtMSADPCM 0.06.

Version 1.15. Tagged as 'HdrSrc-1_15'

Admin:
  This had been deferred pending an update from Pineapple, but I need
    to clear pending updates from pecan's poorly disc.

Version 1.14. Tagged as 'HdrSrc-1_14'

Detail:
  Added IOCTL_PARALLEL_STATUS_BYTE which enters the status byte in the
  least significant byte of the data word of the IOCtl block on return.
Admin:
  This has been tested with the latest parallel port.

Version 1.13. Tagged as 'HdrSrc-1_13'

Detail:
  Should have read:
  Altered the setting for the Modem_RegisterSetOffset from
  &000000 to &800000, to match Lazarus...

Admin:
  This is a correction to the check in log only.
  This version is same as previous (HdrSrc-1_11).

Version 1.12. Tagged as 'HdrSrc-1_12'

Change to modem address offsets for register base and flag to indicate that modem interrupt status register is not present.

Detail:
  Altered the Lazarus setting for the Modem_RegisterSetOffset from
  &000000 to &800000. This can be used to prevent the modem driver accessing
  the memory location that should be occupied by the expansion card header
  (note that the modem registers will actually appear at many addresses
  due to incomplete address decoding). In turn, this means that future
  hardware designs can be arranged so that the modem chip will not
  respond to reads of the expansion card header (by gating nEASCS with
  A23).

  The Venus chip set has no interrupt register, so all interrupts
  from the podule bus are assumed to be from the modem (The modem code
  checks for Modem_InterruptStatusOffset=-1 and doesn't attempt
  to read the Interrupt Status Register if it is true).

Admin:
  Now matches Lazarus settings. Approved by Kevin(tm).

  Unfortunately this change doesn't appear to solve Pace fault 2098
  (operations are very slow due to 6s delay on
  each transaction in some instances. It appears that data is being
  held up in a queue somewhere, or not being sent to the modem).

Version 1.11. Tagged as 'HdrSrc-1_11'

Detail:
    Lazarus is now set to build with the No26bitCode flag set. This allows
32-bit Lazarus builds to be complted without having to alter the HdrSrc files
all the time.
    This does not affect Bethany builds, as these use a branched version of
HdrSrc.

Admin:
    Tested in Lazarus builds.

Version 1.10. Tagged as 'HdrSrc-1_10'

Extra errors added for OS_Hardware.
Hdr:Machine.Machine flag changes for "M_32" build.

Version 1.09. Tagged as 'HdrSrc-1_09'

Detail:

  * New allocation made for South Africa: country number 52.

Admin:

  * Note that the original allocation (territory 130) remains.
    (Alan could give no guarantee that this hasn't been used by third
    parties.)

Version 1.08. Tagged as 'HdrSrc-1_08'