SpindleyQ/honeylisp
1
1
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity
honeylisp/ssc/iigs/u2-debug.fnl

118 lines
3.3 KiB
Plaintext
Raw Normal View History

iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(import-macros {:sss ! : compile} :ssc.macros)
(local {: cmd : response} (require :link.udpdebug))
refactor set!, byte-at, word-at, long-at
2021-09-26 04:34:48 +00:00
; sudo route add -net 172.24.1.0/24 gw 192.168.2.25
iigs: tested, unworking uthernet-ii debug stub
2021-09-25 18:53:18 +00:00
; -VEDRIVE
; CAT,S1
; BLOAD UDPDEBUG.SYSTEM, TSYS, A$2000
; CALL-151
; 2000S
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(local Ssc (require :ssc))
(local ssc (Ssc))
(compile ssc
iigs: tested, unworking uthernet-ii debug stub
2021-09-25 18:53:18 +00:00
(start-symbol boot-8)
(org 0x2000)
(fn boot-8 ()
(asm (clc) (xce) (rep 0x30) ; disable emulation mode, 16-bit index registers
(jsr debug-server-loop)
(sec) (xce))) ; re-enter emulation mode
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(require ssc.stdlib)
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(require ssc.iigs.toolbox)
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(require ssc.iigs.uthernet2)
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(macro out [#(! (WriteLine (far-ref (pstr [$2]))))])
(buffer hexbuf (cstr " "))
(fn printnum (num)
(long! (ref hexbuf) (HexIt num))
(WriteCString (far-ref hexbuf)))
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(global word u2-debug-server-connected 0)
(fn u2-init-debug-server ()
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(out "Starting server")
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(u2-reset)
(u2-udp-server-start 6502)
(set! u2-debug-server-connected 0))
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(buffer u2-debug-buffer 1500)
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
Implement macros, symbol expansion (replaces constants, getters)
2021-09-27 00:32:17 +00:00
(predef-fn u2-debug-server-poll () void far)
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(fn u2-debug-server-cmd-write (msgid)
(let (addr (long-at (ref u2-debug-buffer))
size (word-at (+ (ref u2-debug-buffer) 4)))
(memcpy (+ (far-ref u2-debug-buffer) 6) addr size)
(u2-tx-begin)
(u2-write msgid)
(u2-write [response.ack])
(u2-tx-complete 2)))
(fn u2-debug-server-cmd-read (msgid)
(let (addr (long-at (ref u2-debug-buffer))
size (word-at (+ (ref u2-debug-buffer) 4)))
(u2-tx-begin)
(u2-write msgid)
(u2-write [response.data])
(u2-write-farbuf addr size)
(u2-tx-complete (+ size 2))))
(fn u2-debug-server-cmd-jmp () (asm (jmp ((u2-debug-buffer)))))
(global word u2-debug-server-paused 0)
(fn u2-debug-server-cmd-pause (msgid)
(set! u2-debug-server-paused (not u2-debug-server-paused))
(u2-tx-begin)
(u2-write msgid)
(u2-write [response.ack])
(u2-write u2-debug-server-paused)
(u2-tx-complete 3)
(while u2-debug-server-paused
(u2-debug-server-poll)))
(fn u2-debug-server-cmd-ping (msgid)
(u2-tx-begin)
(u2-write msgid)
(u2-write [response.ack])
(u2-tx-complete 2))
(fn u2-debug-server-cmd (msgid cmd)
(if (= cmd [cmd.write]) (u2-debug-server-cmd-write msgid)
(= cmd [cmd.read]) (u2-debug-server-cmd-read msgid)
(= cmd [cmd.jmp]) (u2-debug-server-cmd-jmp)
(= cmd [cmd.pause]) (u2-debug-server-cmd-pause msgid)
(= cmd [cmd.ping]) (u2-debug-server-cmd-ping msgid)))
iigs: tested, unworking uthernet-ii debug stub
2021-09-25 18:53:18 +00:00
(far-fn u2-debug-server-poll ()
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(let (size (u2-rx-begin))
(when (> size 0)
(let (msgid (u2-read)
cmd (u2-read))
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(printnum msgid) (out "Got message!")
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(u2-read-buf (ref u2-debug-buffer) (- size 2))
(u2-rx-complete)
(when (not u2-debug-server-connected)
(u2-set-tx-dest u2-udp-recv-ip 6502)
(set! u2-debug-server-connected true))
(u2-debug-server-cmd msgid cmd)))))
iigs: tested, unworking uthernet-ii debug stub
2021-09-25 18:53:18 +00:00
(fn debug-server-loop ()
Implement string constants, buffers, macro barriers macro barriers are a hack to say "this form does its own macroexpansion internally" so that we can have state-smart macros that expand to different values depending on if they are being used in code or in a function definition. This seems like a real bad design choice but I can't think of a better one right now! (not quite true: the better design choice is to allow forms to return opgens, and then define a :bytes form that returns the address of the generated thing. :bytes could be understood by :buffer directly. But that's... complicated.)
2021-09-27 03:07:36 +00:00
(TextStartUp)
(IMStartUp)
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
(u2-init-debug-server)
(while true (u2-debug-server-poll)))
)
iigs: tested, unworking uthernet-ii debug stub
2021-09-25 18:53:18 +00:00
(let [Prodos (require :asm.prodos)
prg (ssc:assemble)
disk (Prodos "ProDOS_Blank.dsk")]
(disk:update-volume-header {:name "UDP.DEBUG"})
(disk:add-file "UDPDEBUG.SYSTEM" Prodos.file-type.SYS 0x2000 (. (prg:org 0x2000) :block :bytes))
(disk:write "UdpDebug.dsk"))
iigs: untested uthernet-ii debug stub 65816: support 8-bit immediate mode repl: add re-run and clear buttons ssc: fix returning false from an else clause add byteswap, byte reads / writes start stdlib with memcpy
2021-09-24 02:28:48 +00:00
ssc
Reference in a new issue Copy permalink