Some disk drivers still had internal code for multi-sector I/O. With the read supervisors, this is no longer needed. Removed the extraneous code from these drivers.
- Rearranged HBIOS code to improve colocation of related code.
- Implemented macros to improve control over assembly diagnostic output.
- Improved handling of diagnostic LEDs.
- Improved application boot bank management.
- Implemented application boot for Z280 systems.
- Moved Z280_BNKSEL into proxy.
-
- The post-reset delay of both the IDE and PPIDE drivers has been extended. The SD-IDE adapters need more time to initialize before being ready to behave as proper IDE devices.
- Added support for the FDC section of the Duodyne Disk-IO board.
- Based on reports from Martin R, the FDC detection algorithm has been updated to try reading the FDC MSR register twice to try and get the desired value of 0x80.
- Dan Werner's ESP board for Duodyne is working well, so the default Duodyne config has been changed to automatically detect this board.
Credit to Martin R for reporting that his FDC was not being detected. The delay between FDC status register reads was increased in the detection routine to correct this.
Added a new driver category for DSKY (Display/Keypad) devices. Existing DSKY devices were converted into drivers ICM and PKD. These devices were previously DSKY and DSKYNG.
This removes substantial code duplication and recovers significant space in romldr and dbgmon.
- Modified the Disk Device function call (DIODEVICE) to provide new attributes related to disk size and removability.
- Leveraged API change to allow handling drive assignment differently depending on ATAPI vs. ATA interface.
- Floppy devices are now configured per-device so that each floppy drive can be different (e.g., first floppy is 3.5" and second floppy is 5.25").
- Removed need to use termination signal on floppy sector I/O.
Addresses issue #318
- The DSKY(ng) can display disk I/O activity as the system runs. This capability was implemented very erratically. This update overhauls the mechanism, makes it consistent, and implements it across all disk drivers.
- FD driver will now detect presence of FDC hardware and install if found in all standard configurations
- Does *not* detect type of FDC which must still be set in config
- Added default FD driver MODE to all configs
- Thanks to Alan Cox for the technique!
* added hack to handle tunes
* quiet clean
* added chmod for execution
* suppress warnings
* Multi-boot fixes
* the windows build somehow thinks that these filesystems are cpm3.
* credit and primitive instructions
* Update sd.asm
Cosmetic fix.
* make compile shut up about conditionals
* Add bin2asm for linus and update build to process font files under linix
* fixed quoted double quote bug, added tests
* added tests
* added bin2asm for font file source creation
* Revert linux bin2asm font stuff
* added rule for font source generation
* build fonts
* added directory mapping cache. if the same directory is being hit
as last run, we don't need to rebuild the map. will likely break if
you are running more than one at a time, in that the cache will be
ineffective. also, if the directory contents change, this will also break.
* removed strip. breaks osx
* added directory tag so . isn't matched all over the place
* added real cache validation
* fixed build
* this file is copied from optdsk.lib or optcmd.lib
* install to ../HBIOS
* prerequisite verbosity
* diff soft failure and casefn speedup
* added lzsa
* added lzsa
* removed strip. breaks on osx
* added clobber
* added code to handle multiple platform rom builds with rom size override
* added align and 0x55 hex syntax
* default to hd64180
* added N8 capability
* added SBC_std.rom to default build
* added support for binary diff
* diff fixes
* clean, identical build. font source generator emitted .align. this does not match the windows build
* Upgrade NZCOM to latest
* Misc. Cleanup
* fixed expression parser bug : ~(1|2) returned 0xfe
* added diff build option
* Update Makefile
Makefile enhancement to better handle ncurses library from Bob Dunlop.
* Update sd.asm
Back out hack for uz80as now that Curt fixed it.
* Misc. Cleanup
* UNA Catchup
UNA support was lacking some of the more recent behavior changes. This corrects most of it.
* Add github action for building RomWBW
* Bump Pre-release Version
* Update build.yml
Added "make clean" which will remove temporary files without removing final binary outputs.
* Update Makefile
Build all ROM variants by default in Linux/Mac build.
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update Makefile
* Update for GitHub Build
Case issue in TASM includes showing up in GitHub build. This should correct that.
* Added an gitignore files to exclude generated files
* Removed Tunes/clean.cmd and Tunes/ReadMe.txt - as make clean removes them
* Build.sh: marked as executable
chmod +x Build.sh
* Fix to HBIOS/build.sh
When adding files to rom disk, if files were missing, it would error out.
It appears the intent is to skip non-existing files.
Updated to log out correctly for missing files - and continue operation.
* Update Microsoft NASCOM BASIC.docx
Nascom manual, text version by Jan S (full name unknown)
* Fix issue with Apps/Tune not making
If dest directory does not exist, fails to make Apps
* Create ReadMe.txt
* Update Makefile
* Update Build.sh
* Make .gitignores for Tools/unix more specific
* cpmtools Update
Updated cpmtools applications (Windows only). Removed hack in diskdefs that is no longer required.
* HBIOS Proxy Temp Stack Enhancement
Reuse the bounce buffer area as the temporary stack space required briefly in HBX_INVOKE when transitioning banks. Increases size of temporary stack space to 64 bytes.
* Update ReadMe.txt
* HBIOS - clean up TMPSTK
* Update hbios.asm
Minor cosmetic changes.
* Build Process Updates
Minor udpates to build process to improve consistency between Windows and Mac/Linux builds.
* Update hbios.asm
Add improved interrupt protection to HBIOS PEEK, POKE, and BNKCPY functions.
* hbios - wrap hbx_bnkcpy
* hbios - adjust hbx_peek hbx_poke guards
* Update hbios.asm
Adjusted used of DI/EI for PEEK and POKE to regain a bit of INTSTK space. Added code so that HB_INVBNK can be used as a flag indicating if HBIOS is active, $FF is inactive, anything else means active.
* Add HBIOS MuTex
* Initial Nascom basic ecb-vdu graphics
set and reset for 80x25b screen with 256 character mod
* Finalize Pre-release 34
Final support for FreeRTOS
* Update nascom.asm
Optimization, cleanup, tabs and white spaces
* IDE & PPIDE Cleanup
* Clean up
Make version include files common.
* Update Makefile
* Update Makefile
* Build Test
* Build Test
* Build Fixes
* Update nascom.asm
Cleanup
* Update nascom.asm
Optimization
* hbios - temp stack tweak
* Update hbios.asm
Comments on HBX_BUF usage.
* Update nascom.asm
Optimization
* Update nascom.asm
Setup ECB-VDU build option, remove debug code
* Update nascom.asm
Set default build. update initialization
* Update nascom.asm
Make CLS clear vdu screen
* Update nascom.asm
Fixup top screen line not showing
* Add SC131 Support
Also cleaned up some ReadMe files.
* HBIOS SCZ180 - remove mutex special files
* HBIOS SCZ180 - adjust mutex comment
* Misc. Cleanup
Includes some minor improvements to contents in some disk images.
* Delete FAT.COM
Changing case of FAT.COM extension to lowercase.
* Create FAT.com
Completing change of case in extension of FAT.com.
* Update Makefile
Remove ROM variants that just have the HBIOS MUTEX enabled. Users can easily enable this in a custom build.
* Cleanup
Removed hack from Images Makefile. Fixed use of DEFSERCFG in various places.
* GitHub CI Updates
Adds automation of build and release assets upon release.
* Prerelease 36
General cleanup
* Build Script Cleanups
* Config File Cleanups
* Update RomWBW Architecture
General refresh for v2.9.2
* Update vdu.asm
Removed a hack in VDU driver that has existed for 8 years. :-)
* Fix CONSOLE Constant
Rename CIODEV_CONSOLE constant to CIO_CONSOLE because it is a unit code, not a device type code.
Retabify TastyBasic.
* Minor Bug Fixes
- Disk assignment edge case
- CP/M 3 accidental fall thru
- Cosmetic updates
* Update util.z80
* Documentation Cleanup
* Documentation Update
* Documentation Update
* Documentation Updates
* Documentation Updates
* Create Common.inc
* Documentation Updates
* Documentation Updates
* doc - a few random fixes
* Documentation Cleanup
* Fix IM 0 Build Error in ACIA
* Documentation Updates
* Documentation Cleanup
* Remove OSLDR
The OSLDR application was badly broken and almost impossible to fix with new expanded OS support.
* Bug Fixes
- Init RAM disk at boot under CP/M 3
- Fix ACR activation in TUNE
* FD Motor Timeout
- Made FDC motor timeout smaller and more consistent across different speed CPUs
- Added "boot" messaging to RTC
* Cleanup
* Cleanup
- Fix SuperZAP to work under NZCOM and ZPM3
- Finalize standard config files
* Minor Changes
- Slight change to ZAP configuration
- Added ZSDOS.ZRL to NZCOM image
* ZDE Upgrade
- Upgraded ZDE 1.6 -> 1.6a
* Config File Tuning
* Pre-release for Testing
* cfg - mutex consistent config language
* Bump to Version 3.0
* Update SD Card How-To
Thanks David!
* update ReadMe.md
Remove some odd `\`.
* Update ReadMe.txt
* Update ReadMe.md
* Update Generated Doc Files
* Improve XModem Startup
- Extended startup timeout for XM.COM so that it doesn't timeout so quickly while host is selecing a file to send.
- Updated SD Card How-To from David Reese.
* XModem Timing Refinements
* TMS Driver Z180 Improvements
- TMS driver udpated to insert Z180 I/O waitstates internally so other code can run at full speed.
- Updated How-To documents from David.
- Fixed TUNE app to properly restore Z180 I/O waitstates after manipulating them.
* CLRDIR and ZDE updates
- CLRDIR has been updated by Max Scane for CP/M 3 compatibility.
- A minor issue in the preconfigured ZDE VT100 terminal escape sequences was corrected.
* Fix Auto CRT Console Switch on CP/M 3
* Handle lack of RTC better
DSRTC driver now correctly returns an error if there is no RTC present.
* Minor RTC Updates
* Finalize v3.0.1
Cleanup release for v3.0
* New ROMLDR and INTRTC driver
- Refactored romldr.asm
- Added new periodic timer based RTC driver
* CP/M 3 Date Hack
- Hack to allow INTRTC to increment time without destroying the date
* Update romldr.asm
Work around minor Linux build inconsistency
* Update Apps for New Version
* Revert "Update Apps for New Version"
This reverts commit ad80432252.
* Revert "Update romldr.asm"
This reverts commit 4a9825cd57.
* Revert "CP/M 3 Date Hack"
This reverts commit 153b494e61.
* Revert "New ROMLDR and INTRTC driver"
This reverts commit d9bed4563e.
* Start v3.1 Development
* Update FDISK80.COM
Updated FDISK80 to allow reserving up to 256 slices.
* Update sd.asm
For Z180 CSIO, ensure that xmit is finished, before asserting CS for next transaction.
* Add RC2014 UART, Improve SD protocol fix
- RC2014 and related platforms will autodetect a UART at 0xA0 and 0xA8
- Ensure that CS fully brackets all SD I/O
* ROMLDR Improvements
.com files can now be started from CP/M and size of .com files has been reduced so they always fit.
* Update commit.yml
Run commit build in all branches
* Update commit.yml
Run commit build for master and dev branches
* Improved clock driver auto-detect/fallback
* SIO driver now CTC aware
The SIO driver can now use a CTC (if available) to provide much more flexible baud rate programming.
* CTC driver fine tuning
* Update xmdm125.asm
Fixed a small issue in core XM125 code that caused a file write error message to not be displayed when it should be.
* CF Card compatibility improvement
Older CF Cards did not reset IDE registers to defaults values when reset. Implemented a work around.
* Update ACIA detection
ACIA should no longer be detected if there is also a UART module in the system.
* Handle CTC anomaly
Small update to accommodate CTC behavior that occurs when the CTC trigger is more than half the CTC clock.
* Update acia.asm
Updated ACIA detection to use primary ACIA port instead of phantom port.
* Update acia.asm
Fix bug in ACIA detection.
Thanks Alan!
* MacOS Build Improvement
Build script updated to improve compatibility with MacOS.
Credit to Fredrik Axtelius for this.
* HBIOS Makefile - use env vars for target
Allow build ROM targets to be restricted to just one platform thru use of ENV vars:
ROM_PLATFORM - if defined to a known platform, only this platform is build - defaults to std config
ROM_CONFIG - sets the desired platform config - defaults to std
if the above ENVs are not defined, builds all ROMs
* Added some more gitignores
* Whitespace changes (crlf)
* HBIOS: Force the assembly to fail for vdu drivers if function table count is not correct
* Whitespace: trailing whitespaces
* makefile: updated some make scripts to use when calling subdir makefiles
* linux build: update to Build.sh fix for some platforms
The initialization of the Rom dat file used the pipe (|) operator to build an initial empty file.
But the pipe operator | may sometimes return a non-zero exit code for some linux platforms, if the
the streams are closed before dd has fully processed the stream.
This issue occured on a travis linux ubuntu image.
Solution was to change to redirection.
* Bump version
* Enhance CTC periodic timer
Add ability to use TIMER mode in CTC driver to generate priodic interrupts.
* HBIOS: Added support for sound drivers
New sound driver support with initial support for the SN76489 chip
New build configuration entry:
* SN76489ENABLE
Ports are currently locked in with:
* SN76489_PORT_LEFT .EQU $FC ; PORTS FOR ACCESSING THE SN76489 CHIP (LEFT)
* SN76489_PORT_RIGHT .EQU $F8 ; PORTS FOR ACCESSING THE SN76489 CHIP (LEFT)
* Miscellaneous Cleanup
No functional changes.
Co-authored-by: curt mayer <curt@zen-room.org>
Co-authored-by: Wayne Warthen <wwarthen@gmail.com>
Co-authored-by: ed <linux@maidavale.org>
Co-authored-by: Dean Netherton <dnetherton@dius.com.au>
Co-authored-by: ed <ed@maidavale.org>
Co-authored-by: Phillip Stevens <phillip.stevens@gmail.com>
Co-authored-by: Dean Netherton <dean.netherton@gmail.com>
Floppy I/O was failing for slower CPUs when the timer interrupts were active. Now fixed.
Credit to Jorge Jorge Rodrigues for finding this problem and pointing me in the right direction for the fix.
Supports double-buffered disk i/o to read/write to/from any memory bank specified.
NOTE: Disk I/O will be temporarily slower than normal because all I/O is being forced thru double-buffering for testing.
Wayne Warthen
Dean Netherton
b1ackmai1er
b1ackmai1er