cc65/libsrc/apple2/mli.s

;
; Oliver Schmidt, 30.12.2004
;
; Apple ProDOS 8 MLI
;

        .import         __dos_type

        .include        "apple2.inc"
        .include        "mli.inc"

        .bss

mliparam:.tag   MLI

        .data

callmli:
        ; Store parameters
        sta     call
        stx     mliparam

        ; Check for ProDOS 8
        lda     __dos_type
        beq     oserr

        ; Save random counter
        lda     RNDL
        pha
        lda     RNDH
        pha

        ; Call MLI
        jsr     $BF00           ; MLI call entry point
call:   .byte   $00
        .addr   mliparam

        ; Restore random counter and return
        tax
        pla
        sta     RNDH
        pla
        sta     RNDL
        txa
        rts

        ; Load oserror code and return
oserr:  lda     #$01            ; "Bad system call number"
        sec
        rts
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00			`;`
			`; Oliver Schmidt, 30.12.2004`
			`;`
			`; Apple ProDOS 8 MLI`
			`;`

Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`.import __dos_type`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
Protect random counter against ProDOS. Although documented nowhere (!!!) ProDOS trashes the random counter locations $4E/$4F. Is discovered this because my TCP connections didn't have random local ports. It's a really funny coincidence that David Finnigan discovered only 3 years ago the very same issue because of the very same reason: https://groups.google.com/forum/#!topic/comp.sys.apple2.programmer/1ciep_Oetvo 2018-09-08 21:29:59 +00:00			`.include "apple2.inc"`
Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`.include "mli.inc"`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
			`.bss`

Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`mliparam:.tag MLI`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
			`.data`

			`callmli:`
			`; Store parameters`
Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`sta call`
			`stx mliparam`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
			`; Check for ProDOS 8`
Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`lda __dos_type`
			`beq oserr`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
Protect random counter against ProDOS. Although documented nowhere (!!!) ProDOS trashes the random counter locations $4E/$4F. Is discovered this because my TCP connections didn't have random local ports. It's a really funny coincidence that David Finnigan discovered only 3 years ago the very same issue because of the very same reason: https://groups.google.com/forum/#!topic/comp.sys.apple2.programmer/1ciep_Oetvo 2018-09-08 21:29:59 +00:00			`; Save random counter`
			`lda RNDL`
			`pha`
			`lda RNDH`
			`pha`

			`; Call MLI`
Removed (pretty inconsistently used) tab chars from source code base. 2013-05-09 11:56:54 +00:00			`jsr $BF00 ; MLI call entry point`
Cosmetic changes from Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3477 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-04-20 16:52:10 +00:00			`call: .byte $00`
			`.addr mliparam`
Protect random counter against ProDOS. Although documented nowhere (!!!) ProDOS trashes the random counter locations $4E/$4F. Is discovered this because my TCP connections didn't have random local ports. It's a really funny coincidence that David Finnigan discovered only 3 years ago the very same issue because of the very same reason: https://groups.google.com/forum/#!topic/comp.sys.apple2.programmer/1ciep_Oetvo 2018-09-08 21:29:59 +00:00
			`; Restore random counter and return`
			`tax`
			`pla`
			`sta RNDH`
			`pla`
			`sta RNDL`
			`txa`
Much better approach for ProDOS 8 IO-buffer custom implementations - the first one was actually kind of stupid. git-svn-id: svn://svn.cc65.org/cc65/trunk@4155 b7a2c559-68d2-44c3-8de9-860c34a00d81 2009-09-11 19:54:29 +00:00			`rts`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00
			`; Load oserror code and return`
Removed (pretty inconsistently used) tab chars from source code base. 2013-05-09 11:56:54 +00:00			`oserr: lda #$01 ; "Bad system call number"`
dio implementation by Oliver Schmidt git-svn-id: svn://svn.cc65.org/cc65/trunk@3432 b7a2c559-68d2-44c3-8de9-860c34a00d81 2005-03-30 18:25:04 +00:00			`sec`
ProDOS 8 requires an page-aligned 1kB IO-buffer for every open file. The implementation using posix_memalign() is universally correct but quite expensive. There I added support for overriding this implementation with maybe limited (i.e. to one open file) but cheaper custom implementations. git-svn-id: svn://svn.cc65.org/cc65/trunk@4142 b7a2c559-68d2-44c3-8de9-860c34a00d81 2009-09-10 09:59:26 +00:00			`rts`