GitLab

When no drives are detected (but some are configured) CDFSSoftATAPI goes dormant by setting its private word to 0. However, when the module finalises (eg. an RMReinit) that NULL value is then used to call CD_Unregister, and attempting to free the same NULL value with OS_Module.
Also replaced magic numbers from Hdr:ModHand and Hdr:OsBytes.

Version 1.34. Tagged as 'ATAPI-1_34'

Disassembly unchanged apart from the single register STM's which now use STR ith writeback.

Version 1.33. Tagged as 'ATAPI-1_33'

Changed debugging to run with DebugIt prior to possible code changes.  No actual code changes, so module version not changed.

Detail:
Admin:  built in test tungsten rom. works


Version 1.32. Not tagged

Now responds to "Driver in use" errors returned from ADFS_ATAPIOp by retrying up to the specified timeout (or 10 seconds for internally generated

SCSI_Ops). If the timeout is reached, the CD_SCSIUserOp entry point will
return a SCSI-compatible "ATAPI driver reservation error" error pointer;
other entry points pass "Driver in use" through unchanged.
Also added GPA generation capability, for what it's worth.

Version 1.32. Tagged as 'ATAPI-1_32'

Speed setting tweaked one more time. Now asks for 176.4 * speed - this works a bit better on some drives (eg Sony) which round down. Nominal basic CD rate is 176400 bytes/second.

Version 1.31. Tagged as 'ATAPI-1_31'

Detail:
  * Byte sex swap fix applied to the code to set the drive speed
  * SCSI-3 spec says that ATAPI packets have to be padded with 0 bytes;
    routines changed accordingly (though this does mean we now have to do
    a tedious copy onto the stack for every operation)
  * All disc errors returned from ADFS_ATAPIOp were accidentally ignored!
  * Internal disc number wasn't being shifted left by 15 bits before being
    used to do a REQUEST SENSE command packet, resulting in it doing a
    reset of any slave devices instead of sending a PACKET command
  * TARGETERROR error numbers weren't being correctly constructed in
    ATAPI_Op (it was using the error pointer instead of the sense key)
  * Worked round another assumption of flag preservation in CD_ SWIs
  * Fixed pathname of messages file to include $
Admin:
  Tested on Tungsten.

Version 1.30. Tagged as 'ATAPI-1_30'

Detail:
  * Removed IOMD dependency - reimplemented to use ADFS_ATAPIOp and
    ADFS_IDEDeviceInfo
  * Supports 4 drives (previously 2)
  * Support for SCSIUserOp entry included
Admin:
  Requires ADFS 3.34 (or intermediate version after 3.33).

Version 1.29. Tagged as 'ATAPI-1_29'

Detail:
  * Moved resources here from CDFSResources - allows tokenisation
  * Uses objasm, shared makefiles and global headers
  * 32-bit compatible
Admin:
  Not tested. Still IOC/IOMD-centric.

Version 1.28. Tagged as 'ATAPI-1_28'

Detail:
  This module had sets of defines for what options to have in what
   version of the module, instead of just using different versions out
   of a source control system...very odd.  Changed code to say
   "if version is over 1.24, use these standard flags".  Also removed
   check for a "too high" version number.
Admin:
  Built.


Version 1.27. Tagged as 'ATAPI-1_27'

Detail:
  Merged Ursula branch, only change was removal of dynamic dependencies
   from Makefile.

  Moved to srccommit usage, changed some occurences of "Version" to
   "ModuleVersion", as defined in VersionASM. Removed obsolete "Version"
   file.
Admin:
  Added stripdepend call to clean rule
  Additionally added a "GET Hdr:Machine.<Machine>" into s.Main to
   make it assemble correctly.


Version 1.26. Tagged as 'ATAPI-1_26'