GitLab

Now, for each SWI call there is a definition of any pertinent structures it asks for and definitions of any bitfields within those flags. This avoids the need for clients to endlessly redefine these locally (in practice it looks like sections of FileCore were simply copy and pasted into clients RAMFS/SCSIFS/ADFS). Delete private definitions.
This binary was carefully checked to be identical since so many locations were changed.
Then, the following additional changes:
* InitDieSvc line 74, the floppy config is extracted using a mask and shift rather than reaching up the stack
* Identify lin 1254, the superfluous instruction marked as such deleted
* FileCore15 line 762 recoded the check for background op to not need the bit number defined any more
* FileCore00 moved the label 'anull' to be word aligned guaranteed
The duff pointer marker ('nowt' = &40000000) is no longer used to mark territory translation tables as invalid as that address is now quite reasonable. -1 is used instead.

Version 3.45. Tagged as 'FileCore-3_45'

When failed to identify, and the last operation was format, the DiscSize2 field was not being copied (instead DiscSize is copied twice).
Fix up some warnings from the assembler.
Tested briefly with a RAM disc, still works.

Version 3.44. Tagged as 'FileCore-3_44'

R8 was being used to remember the C bit, by copying from the PC. Swapped to use RRX instead.

Version 3.43. Tagged as 'FileCore-3_43'

Entry macro renamed as SemEntry to avoid conflict with Hdr:Proc.
Other macro bit rot fixed up.
Tidy up switches.
DebugFx switch is the only one that doesn't work.
Non debug binary same as 3.41.

Version 3.42. Tagged as 'FileCore-3_42'

R5 was being reused as the loop exit instead of R9.
Synced code with *CHECKMAP a bit since they have the same heritage.
Untangle the spaghetti around TermCommon and UnTermCommon.
In the module init, skip a store when the flags don't change.
Tested with a RAM disc filled with autogenerated nested directories and a loop blindly trying to defect everything - works fine now. Ticket #114.

Version 3.41. Tagged as 'FileCore-3_41'

Detail:
  The Density field in the disc record is a byte at a non-word-aligned address
  but it was being loaded here with an LDR instruction. There is actually a
  pre-existing bug here, since the code at lines 904-905 would only have been
  executed on earlier CPUs if SectorSize, SecsPerTrk and Heads were all zero
  too. This would correspond to an implausibly small hard disc, so arguably
  this part of the branch should perhaps be removed altogether?
Admin:
  Untested at ROOL.
  Submission from Willi Theiß [comments above are my own - BA]

Version 3.40. Tagged as 'FileCore-3_40'

Detail:
  s/FileCore, s/FileCore00, s/InitDieSvc - Try using OS_ReadSysInfo 6 to get IRQsema & CannotReset locations before falling back on legacy values. Store results in module workspace.
  s/FileCore05 - Commented out unused CheckEscape routine to avoid having to update it
  s/FileCore15 - Debugging code updated to use OS_ReadSysInfo 6 to fetch IRQsema ptr. Can't always rely on workspace version since workspace might not be set up yet.
  s/FileCore25, s/FileCore30, s/FileCore80 - Use IRQsema & CannotReset pointers from workspace
Admin:
  Tested on rev A2 BB-xM


Version 3.39. Tagged as 'FileCore-3_39'

Version 3.38. Tagged as 'FileCore-3_38'

Combat slow USB devices on boot by making OS_FSControl 15 auto-retry when SCSIFS "drive empty" errors are generated

Detail:
  s/FileCore60 - DoOsFunBootup now detects when a SCSIFS "drive empty" error is returned by IdentifyCurrentDisc. Instead of returning with an error, the boot operation will instead be retried, to give slow/unlucky USB devices a chance to finish initialising. Escape can be used to cancel the retry loop, just in case.
  Resources/UK/Messages - Added new 'RetryBoot' message which is shown when the retry loop starts
Admin:
  Tested on rev A2 BB-xM. Machine now boots correctly in situations where the required USB device hasn't initialised by the time OS_FSControl 15 is called.
  New message token (RetryBoot) needs translating.


Version 3.37. Tagged as 'FileCore-3_37'

Detail:
  FindDiscRec was erroneously using LDR to load DiscsDrv when it should have used LDRB. Apart from the potential for disc record priorities to be incorrectly evaluated, this was also causing an abort when alignment exceptions were turned on.
Admin:
  Tested on rev C2 beagleboard. FAT formatted USB drives now seem to work OK.


Version 3.36. Tagged as 'FileCore-3_36'

Detail:
  The code which protects against hash table overflows in the CompactZone routine has been broken ever since BigMaps was introduced. With BigMaps each hash table entry is three bytes, not two; but the overflow fixup code which restores the word following the hash table to its original value wasn't updated to take this into account, and so was restoring the data to the wrong position.
  Presumably this isn't a critical problem since it's been there so long, but since it causes an alignment exception when NoUnaligned is in use it's now been fixed.
Admin:
  Tested on rev C2 beagleboard


Version 3.35. Tagged as 'FileCore-3_35'

Detail:
  Many of these are unaligned LDRs where only bits 0-7 of the result are used
  (tested against a bitmask or used as an input to the barrel shifter) but
  they have all been moved to byte operations because they are faster on
  ARMv6 and won't cause false positives if unaligned aborts are enabled.

  s.BigDirCode line 104: not wrong as such, but changed to use aligned LDR
    for speed on ARMv6.
  s.BigDirCode line 1727: LDR of an unaligned 1-byte variable. Only used as
    input to shifter.
  s.Commands line 129: STR to an unaligned 1-byte variable. Harmless because
    all overlapping variables are initialised later on.
  s.Commands line 539: LDR of an unaligned 1-byte variable. Only used for
    testing bit 6.
  s.FileCore31 line 1359: LDR of an unaligned 1-byte variable. Only used as
    input to shifter.
  s.FileCore32 lines 1483 and 1825: LDR of an unaligned 1-byte variable. Only
    used as input to shifter.
  s.FormSWIs line 1403: bugfix: incorrect load of 16-bit zone_spare field of
    disc record - uses lowsector and nzones fields instead! This is used to
    terminate the search of free space fragments within the allocation bytes
    of the current map block for the one containing a bad block found during
    formatting. Would manifest itself as a "Can't map defect out" error.
  s.GenSWIs lines 1182,1195: bugfix: LDR of 1-byte variables. Effect of this
    was that range-checking of the drive number passed to FileCore_MiscOp 5
    (eject) wasn't performed.
  Added ENTRY directive to permit building of GPA debug listing.
Admin:
  Builds, but untested.

Version 3.34. Tagged as 'FileCore-3_34'

Detail:
  Uses suffixed file extensions for compatiblity with both objasm and asasm.
Admin:
  Supplied by Peter Naulls, tested at ROOL

Version 3.33. Not tagged

Detail:
  * ARMv6+ builds use unaligned halfword and word loads and stores (unless
    pre-ARMv6 compatibility is required or NoUnaligned is set)
  * In many cases, one unnecessary BIC instruction has been removed from
    pre-ARMv6 builds
  * ARMv5+ builds (so Tungsten and OMAP) make use of CLZ instruction when
    parsing fragment blocks
Admin:
  Built and included in an OMAP3 ROM. But be warned, this has not received
  the rigorous testing normally required of filesystem code, use at your
  own risk.

Version 3.33. Tagged as 'FileCore-3_33'

Skip "Disc drive empty" error on shutdown - if the drive's empty there's not much you can do about it.

Dir cache flushing fixed.

Version 3.32. Tagged as 'FileCore-3_32'

* Can use HAL for FIQ+timer manipulations.
* FileCore_DiscOp64 added (as user API - currently just converted to the
  standard byte or sector DiscOp call to the underlying module). See
  Doc.32bitAPIs for full details.
* New error-passing scheme to modules added (for full 32-bit addressing).
* FileCore_Features added (indicates new error scheme).
* OS_FSControl 49 (read 32-bit free space) now reports sensible clamped
  values for large discs, rather than values modulo 2^32.
* Accessing the end of files near 2GB in size could cause lock-ups in
  background transfers.
* F(+) and G(+) disc formats adjusted to have a little more skew to
  help the rather poor Iyonix floppy controller.

Version 3.31. Tagged as 'FileCore-3_31'

Version 3.30. Tagged as 'FileCore-3_30'

  Updated to build using objasm instead of aasm.
  Sources changed to be objasm-compatible.
Admin:
  Requires Library 0.72 or later.
  Requires BuildSys 3.09 or later.
  Requires Env 0.65 or later.

Version 3.29. Tagged as 'FileCore-3_29'

Version 3.28. Tagged as 'FileCore-3_28'

Detail:

  * No longer reliant on STB flag.  Now both Desktop and Embedded builds
    are the same (which, of course, they should be).

Admin:

  * StripDepnd support moved to MkClean file.

  * The "What'sHere" file has been renamed to "ReadMe", since the existence
    of the single-quote causes problems with some Unix commands (notably
    'xargs').

  * Built and tested in a Lazarus 32-bit build.

Version 3.27. Tagged as 'FileCore-3_27'

Detail:
  Scatter list wrap condition updated to allow top-bit-set memory accesses.
Admin:
  Not tested - any issues related to the 32-bit conversion have not been
  addressed. I'd suggest that you use the branched version mentioned above,
  except in test builds.

Version 3.26. Tagged as 'FileCore-3_26'

Detail:
  Flag preservation added to 7 routines that require it.  Docs.32-bit
    updated to remove the names of the functions that have been made
    to preserve flags again.
Admin:
  This version needs more testing, but at least this version does not
    appear to stiff the machine, corrupt discs, go bang under light
    testing with ADFS and RAMFS.
  26-bit and 32-bit builds slightly tested.
  Do not use in products - Do use in test builds.

Version 3.25. Tagged as 'FileCore-3_25'

Detail:
  Some stack offsets were wrong - now fixed.
  More PSR preservation added.
Admin:
  26-bit builds of this module still stiff the machine when ADFS is
    used.
  Do not use - except for testing.

Version 3.24. Tagged as 'FileCore-3_24'

Version 3.23. Tagged as 'FileCore-3_23'

Detail:

  * The latest merge appeared to have missed one of the "Module_Version"
    symbol changes.  It failed to assemble:

      s.FileCore40: Serious error  Undefined symbol

    "Version" changed to "Module_Version" to correct this, in line with
    the other related changes.

Admin:

  * Untested; to be built into next Lazarus development build.

Version 3.22. Tagged as 'FileCore-3_22'

Detail:
  This module represents the latest version of FileCore incorporating
    both the Ursula changes and the ROL changes as we have them to date.
  Changes from ROL-FileCore-3_21 are basically the stripping of trailling
    spaces and the change of Version to Module_Version in the conditionals.
    No other changes.
Admin:
  ROL branch tagged ROL_merge
  Ursula branch tagged Ursula_merge.
  Untested, but essentially the same as ROL-FileCore-3_21, so no problems
    expected.

Version 3.21. Tagged as 'FileCore-3_21'

Version 3.01. Tagged as 'FileCore-3_01'

Version 3.00. Tagged as 'FileCore-3_00'

Makefile changed to use LocalRes$Path.
Changed to use srccommit.
StrongARM flag now set to :LNOT:STB

Version 2.99. Tagged as 'FileCore-2_99'