| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An MBC5 cart has no RTC, so SNTP is the only route to a wall clock. ntpdate
does one 48-byte UDP exchange on port 123 and hands the stamp to the kernel;
SYS_SETTIME keeps it with the tick it arrived at, and SYS_GETTIME adds the
elapsed ticks back, so date(1) is just formatting.
/# ntpdate
ntpdate: clock set from pool.ntp.org
/# date
Sep 24 21:01:23 UTC 2026 (host said 21:01:26)
The clock does not survive a power cycle - which is exactly what /rc is for:
`ntpdate` there and the machine boots knowing the time.
All of it is bounded subtraction, no long division or shifts: SDCC pulls those
from sm83.lib, whose modules link into the $A000 RAM window. The month tables
are string literals for the same family of reason - an initialized array goes
to _INITIALIZED at $A000 and the image overflows its bank ("buffer too small"
out of makebin), which is how this first failed to link.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
/# ls -l /bin > b
/# cut -d \s -f 1 b | uniq -c
52 -
- touch FILE...: gbfs has no timestamps (an MBC5 cart has no RTC), so touch
does the half that means something here - make an empty file. O_APPEND is
exactly right: it creates, and it does not truncate what is already there.
- tr [-d] SET1 [SET2]: ranges (a-z), -d to delete, and a short SET2 padded
with its last character, like tr(1). ASCII only - a 128-entry map.
- rev, uniq [-c]: one line each, from a file or stdin like the other filters.
Two things the tests caught. A bare "-" is a SET, not a flag, so `tr . -`
has to skip flag-matching only when a letter follows the dash. And uniq
compared an empty "line" against the last run at EOF and flushed a phantom
blank one - a stray "1" in uniq -c output.
tr and cut also take \s \t \n \r escapes, because the shell has no quoting:
`-d ' '` arrives as three tokens, so a space or tab delimiter was simply
untypeable. Shell quoting would be the better fix; this makes the tools
usable today without touching the parser.
Banks 49-53; 10 program slots left in the 1 MB ROM.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The commands only existed in the kernel's program table, so the only way to
find out what the machine could do was to already know (or read the source).
They are now a directory in the tree:
/# ls / /# xxd -l 16 /bin/uname
bin 0000: cd 6b 43 06 00 0e 00 f7 .kC.....
readme 0008: 18 fe 47 0e 03 f7 c9 21 ..G....!
/# ls -l /bin | head -2
- 1 16384 worker
/bin is synthetic, not disk: dir_find resolves "bin" in the root to BIN_INO
and names inside it through the NameTable, so `ls /bin` lists exactly what the
shell can run - one table, one truth, no second list to keep in sync. The root
listing carries "bin" as a synthetic first entry so it is discoverable by
walking the tree.
A program image is inode BIN_FILE|progid: it stats as a 16 KiB file (a program
IS its ROM bank) and reads through rom_getb, which maps that bank, takes the
byte, and maps the CALLER'S text bank back before returning - the caller is
itself executing from $4000, so leaving the wrong bank mapped would return
into another program's code. Verified byte-for-byte against the host's view of
the blob. The tree is read-only: create/remove/write inside /bin all fail.
xxd grows -l (stop after N bytes), because a program image is 2048 dump lines
- more than a pipe's temp file can hold, let alone the screen.
Two traps worth remembering, both found the hard way here: `and BIN_FILE`
destroys A, so the inode must be reloaded before inode_ptr (missing it once
made ls -l stat inode 0 and call every file an empty non-directory); and ls
now copies its directory argument to the stack first, because stat()'s request
block is a static living at $A000 - on top of the command line it was reading.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Eight bytes per line rather than the usual sixteen: offset + 8 hex pairs +
the ASCII column is 38 of the terminal's 40 columns, where 16 would wrap
every line and make the dump unreadable. -c overrides it.
No -s (skip): there is no seek syscall, so it streams from the start - the
same reason tail(1) has to keep a ring.
/# xxd h
0000: 48 65 6c 6c 6f 20 67 62 Hello gb
0008: 6f 73 0d 0a os..
blk(1) could already peek at raw disk blocks; nothing could look inside a
file. Verified against known bytes, from stdin, with -c, through a pipe, and
on a missing file.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Filling in the core utilities the box was missing:
- ls -l: type, links and size, now that stat() exists. The screen is 40
columns, so it is "d 1 256 name", not the GNU sprawl. Sizes were
previously unknowable except one file at a time with wc -c.
- df: blocks and inodes, used and free, on the 32 KiB cart disk.
- mv: no rename syscall, so it is cp + rm - and the copy has to finish before
the source goes, or a failure loses the file.
- tail [-n N]: no seek syscall, so it streams the input through a ring of N
lines rather than jumping to the end. N is capped by that ring.
- tee: the missing half of a pipeline - `cat big | tee copy | wc -l`.
- sleep: the thing every script needs; the kernel counts 1/64 s ticks.
- help: lists every command the shell knows, walked out of the kernel's own
name table. The box previously had no way to tell you what it could do.
- poweroff: the only safe way to stop a gbos that has data on it. The cart's
battery RAM is written when the emulator exits cleanly, so killing it loses
the whole filesystem - it comes back "formatted". Found while testing /rc.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
/# ircd &
host$ irssi -c 10.0.0.2 /join #gbos
other GB$ irc 10.0.0.2 gb2 /join #gbos
Implements the subset a real client needs to get in and talk: NICK, USER,
PING/PONG, JOIN, PART, PRIVMSG/NOTICE (to a channel or to a nick), NAMES and
QUIT, with the 001/004/375/372/376 numerics clients wait for on registration,
353/366 on join, and 421 for anything else. One channel per client keeps the
bookkeeping in fixed-size tables; this is a handheld with 8 KiB of task RAM.
MAXCL is 3, bounded by the kernel socket table (listener + client each take a
slot, the rest left for other programs). A fourth client is refused and the
daemon keeps serving - which is how the ~30s close() stall turned up.
Every event is logged to the LCD, so the Game Boy shows its own server
traffic: "+ alice", "alice joined", "<alice> hello".
Verified with real clients over the hub: two and three simultaneous hosts
registering, joining, channel relay (sender not echoed to itself), private
messages by nick, PING/PONG, unknown-command numerics, QUIT propagation, the
overflow refusal, and a second Game Boy running gbos's own irc(1) in the same
channel as a laptop client - messages crossing both ways between them.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
/# httpd &
host$ curl http://10.0.0.2/ -> a listing of the cart RAM
host$ curl http://10.0.0.2/readme -> the file, from battery-backed SRAM
GET only. "/" serves /index.html if present, else a generated listing; any
directory lists with every entry linked, so you can click around a 1997
handheld's disk from a browser. Content type from the extension. No
Content-Length: HTTP/1.0 + Connection: close delimits the body by closing,
which suits a stack with one in-flight segment.
All output goes through one 200-byte buffer, so each net_send is a full
segment instead of one tiny segment per string.
It drains the whole request, not just the request line. That is not
politeness: leaving the rest unread means those segments are never
acknowledged, so the client retransmits headers for a minute, never sends
its FIN, and its dead connection keeps arriving at port 80 long after we
re-listen. One curl with 1.4 KB of headers stalled every following request
for 30-120s.
It deliberately does NOT read the keyboard either. Backgrounded - the useful
way to run it - every key it polled would be a key stolen from the shell's
prompt. It is a daemon: kill(1) stops it, and the kernel reclaims its socket.
Verified with curl: index, byte-exact file bodies, nested paths, directory
listings, 404, 501, a 1.4 KB request spanning segments, browser user-agents,
three parallel clients (0.1/1.1/2.1s), the shell staying usable alongside it,
and six kill/restart cycles - more than there are sockets.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The shell has had pipes for a while with nothing to pipe *through*, and
an 18-row screen with no way to stop output scrolling off it.
- grep [-vin] PAT [file]: substring match (no regex), stdin or a file,
grep(1)'s exit convention. The other end of the pipe, at last.
- more [file]: pages 17 lines at a time, --More-- bar in white-on-blue,
any key pages, q quits. Counts *screen* rows, so wrapped lines pay
their real cost. Keys come from pollin/pollcon, not stdin, so
`cat big | more` still has a keyboard.
- cp SRC DST: the fs has rm/mkdir but no way to duplicate a file. mv is
this plus rm (there is no rename). Rejects SRC == DST, which would
otherwise truncate the source via O_WRITE before reading a byte.
- nc [-s] HOST PORT: raw TCP. Interactive (START sends the line with a
CRLF, /q quits) or -s to pump stdin and print the reply with its own
wall-clock idle timeout, because the kernel's blocking recv counts
pumps, not seconds. Generalizes what chat/irc hardcode:
`nc towel.blinkenlights.nl 23` works.
sys_pollcon now pumps the link before draining the console ring. The
ring is filled *by* net_pump, so a program that never touches the
network polled a ring nothing would ever fill: host-injected keys
(gbtype/gbdemo) hung `more` forever, and the Game Boy stopped answering
pings for as long as it ran. KGetc already pumps for this reason.
New tools take banks 35-38. Locals over statics in all four: a program's
_DATA starts at $A000, where the shell leaves the command line.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
Same setup as sl0pboy: git-clang-format scopes formatting to the lines
each commit touches, re-adds the formatted files, and refuses
partially-staged C files. Style file matches the house style (4-space,
attached braces, 100 cols, short cases/ifs inline, no include sorting -
order is load-bearing for SDCC userland). Enable once:
git config core.hooksPath hooks
|
| |
|
|
|
|
|
|
|
| |
README is now pitch -> proof (real capture gif) -> quick start -> docs
pointers, with the network stack and gbhub/gbjoin/gbtype getting
explicit billing. The full technical breakdown (memory model, syscall
ABI, fork/exec/context switch, FS layout, toolchain war stories) moves
to docs/internals.md, plus previously-undocumented sections on the
network stack and the terminal/OSK.
|
| |
|
|
|
|
|
| |
Userland sources live in usr/ (fits the Unix theme better than 'c').
SDCC output .bin blobs land in build/usr/ with the other build artifacts
instead of littering the source dir; programs.asm INCBINs them from there.
Byte-identical ROM.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- directories generalized: dir_find/dir_add/dir_remove take any dir inode;
every dir carries '.'/'..'. Path resolution (resolve / resolve_parent) walks
absolute or cwd-relative paths; open/mkdir/chdir/remove/opendir take paths.
- per-process cwd: PROC_CWD in the PCB (root by default, inherited on fork);
sys_chdir sets it, sys_opendir points ls at any directory.
- shell: 'cd' builtin + cwd-aware prompt; 'mkdir'/'ls <dir>' tools; 'exit'/'quit'
and EOF call poweroff() (clean shutdown via the $ED opcode -> reliable save).
- three HL/buffer-clobber bugs fixed along the way: resolve didn't preserve the
path cursor across dir_find; cur_cwd clobbered HL; resolve_parent set the final
name before resolve() overwrote wFsNameBuf (now copied after).
- verified: nested mkdir/cd/save, multi-level paths (cat docs/sub/b), and the
whole tree persists across a reboot.
- README: document directories + the poweroff opcode.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- fs.asm: a real block filesystem on the persistent block device - superblock,
block/inode bitmaps, a 32-entry inode table (type, size, 8 direct block
pointers -> files up to 2 KiB), and a root directory of 16-byte entries.
Metadata cached in WRAMX; one-block data cache streams file/dir blocks.
- same syscall interface (open/getb/putb/list/remove) -> tools unchanged;
ls now enumerates until flist() runs out (16 files, was hardcoded 8).
- old 8-slot WRAM FS removed; cart RAM partitioned: process banks 0-7, disk 8-11.
- two register-clobber bugs fixed: alloc_block/alloc_inode returned the
bitmap-block number (write_bitmap clobbers C); db_use loaded the wrong block
on a transition (db_flush->write_block clobbers C) -> multi-block files broke.
- verified: 10+ files, multi-block (300-byte) files, delete, and persistence
across reboots. Debug 'blk' disk tool retained (fill/peek).
- README: document the block filesystem.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
- blk.asm: read_block/write_block move 256-byte blocks between battery-backed
cart-RAM banks (8-11, the 'disk') and a WRAM bounce buffer. Banking is hidden
in those two routines; they inline the 'restore my bank' step so they never
touch the stack while the disk bank is mapped over the process's $A000 window
(the stack lives there too -- a call/ret would rug-pull it).
- format-on-first-boot: superblock magic in block 0; otherwise the disk persists.
- cart RAM bumped to 128 KiB (-r 4); process banks 0-7, disk banks 8-11.
- debug tool 'blk fill/peek' + syscalls to validate round-trip and persistence.
- verified: round-trip incl. cross-bank (block 100 -> bank 11); block survives a
reboot via .sav; magic intact (no reformat on 2nd boot).
- old WRAM 8-slot FS still backs the tools until stages 2-3.
|
| |
|
|
|
|
|
|
|
|
|
| |
- sys_reap: nohang reap of a finished zombie child (-> pid or 0).
- shell: 'cmd &' launches in the background (prints [pid], no wait); reaps
finished jobs at the prompt ([pid done]). Foreground now waits for its own
child specifically (loops wait(), reporting bg completions meanwhile) so kill
reports the right pid.
- spin: a process that yields forever - a target for ps/kill.
- verified: spin &; ps shows it; kill <pid> removes just that one; immortal
init; self-finishing bg worker reaped with [pid done].
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
- kernel I/O routing: PCB gains PROC_STDIN/PROC_STDOUT (default console $FF),
inherited across fork. KGetc/KPutc route read/write/putc to the console or a
file fd. New syscalls: putc(16), setin(17), setout(18), lookup(19).
read/write now go through the routing; putc/puts/nl use SYS_PUTC.
- sys_lookup + NameTable move command-name resolution into the kernel.
- shell rewritten in C (c/sh.c): tokenizes a line, parses > / < / |, and drives
fork+setin/setout+exec+wait. Pipes run as 'a > __pipe ; b < __pipe' (temp
file). asm shell + cmdtab removed; StrEqual/SkipName kept for sys_lookup.
- libc: fork/exec/wait/setin/setout/lookup wrappers.
- verified: echo>file, cat file, wc<file, cat readme|wc -l, echo ...|wc.
|
| |
|
|
|
|
|
|
|
|
|
| |
- fs.asm: 8-slot RAM filesystem in WRAMX ($D000-$DFFF), name[8]+len[2]+data[502];
syscalls open/close/getb/putb/list/remove, seeded with a 'readme' at boot.
- libc: open/close/fgetc/fputc/flist/fremove wrappers + O_READ/O_WRITE/NOFD.
- tools: ls, save (stdin->file, one line), rm; cat/wc/head now take a file arg.
- fix: syscall dispatch clobbers A, so putb takes its byte in E (was reading the
handler's low address byte, 0xDC); ls NUL-terminates 8-char names.
- verified: save/cat/ls/rm cycle; wc readme -> '2 7 38'; head -n 1 readme.
- README: document the filesystem + the A-clobber ABI note.
|
| |
|
|
|
|
|
|
|
|
| |
- c/libc.c: shared C helpers linked into every program - putu (print decimal),
atou (parse decimal), argv_parse (tokenize getargs() into argc/argv).
- wc: count lines/words/chars of stdin. head [n]: first n lines (drains rest to
EOF). args: argc/argv demo.
- pid now uses libc putu; build.sh links libc.c; Makefile CBLOBS + libc dep.
- README: document the new tools + argv_parse.
- verified: 'args one two three' -> argc=3/argv[..]; wc '2 3 16'; head 2.
|
| |
|
|
|
|
|
|
|
|
|
| |
- libc: add putc, puts, strlen, getargs (args string), EOF (=4) for readc.
- args: shell splits the command line at the first space ("cmd\0args\0" at
$A000, inherited by the child via fork); getargs() returns the arg string.
- shell: exit on Ctrl-D/EOF; $ prompt over a clean read loop.
- tools (c/): echo (argv), cat (stdin->stdout to EOF), uname, pid (decimal
print), true, false. Each built to a ROM-bank blob and registered.
- Makefile: CBLOBS list builds all C programs.
- README: document libc + the tool set.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
- c/{gbos.h,crt0.s,libc.s,build.sh}: SDCC sm83 -> ROM-bank blob toolchain.
libc syscall wrappers save/restore BC/DE/HL around rst $30 (trap clobbers
them; SDCC expects them preserved).
- build via SDCC native asxxxx path (sdasgb/sdldgb), crt0 linked first so
_start is the $4000 entry; blob INCBIN'd into a ROM bank.
- chello.c: prints a message + getpid() -> runs as 'chello' shell command.
- Makefile: auto-build c/*.bin, track as a dep of programs.o; gitignore blobs.
- README: document the C toolchain + the SDCC --asm=rgbds codegen bug that
forced the native-toolchain approach.
- verified: '$ chello' -> 'hello from C on gbos!' / 'my pid is 2'.
|
| |
|
|
|
|
|
|
|
|
|
| |
- root cause of the read/echo corruption: SyscallTrap uses HL to index the
dispatch table, so HL is clobbered across a syscall; the shell then wrote its
buffer pointer's garbage into the $0000-$7FFF MBC register region, corrupting
the mapper. fix: userland saves HL around sys_read (push/pop).
- shell now runs commands end to end: $ hello -> hello world!, $ worker ->
worker!, $ foo -> ?
- README: document that syscalls clobber BC/DE/HL (return in A, B for wait);
add read + shell to the syscall table.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
- exit(): zombie + status, reparent orphans to init, wake blocked parent,
schedule away forever (resources reclaimed by the reaper)
- wait(): scan for a zombie child; reap (free cart-RAM bank + PCB slot) and
return pid/code; block PS_BLOCKED + yield if children still live; $FF if none
- cart-RAM banks are now a free list (wBankUsed bitmap): AllocRamBank/FreeRamBank
- scheduler: FindNextReady returns carry when nothing runnable; yield idles
instead of blindly picking slot 0
- helpers: FindPcbByPid, ReparentToInit
- demo: init forks 3 workers -> exec -> exit(pid) -> wait/reap => wwwR2R3R4!
verified bank recycling with 6 workers over 3 banks (w2w3w4w5w6w7)
- README: document the lifecycle + syscall status
|
| |
|
|
|
|
|
|
|
|
| |
- sys_exec(B=program id): look up ProgramTable, point PROC_ROMB at the program's
ROM text bank, map it into $4000-$7FFF, reset stack, jp to entry (no return)
- programs.asm: two demo programs (ping/pong) ORG'd at the SAME $4000 in banks
2 and 3 - proves execution is driven purely by PROC_ROMB
- demo: init forks; parent execs PROG_PING (bank2, '1'), child execs PROG_PONG
(bank3, '2') -> '1212...', 2000/2000 balanced, zero garbage
- README: document exec + text-in-ROM model
|
| |
|
|
|
|
|
|
|
|
| |
- sys_fork: switch to a kernel stack, bump-allocate a cart-RAM bank, copy the
parent's 8 KiB bank via a WRAM bounce buffer, plant a switch-in frame so the
child returns 0 to the post-fork PC, fill the child PCB (shared ROM text)
- add AllocRamBank (bump allocator) + FindFreeSlot helpers
- demo: init forks; parent prints P, child prints C (PCPC...); nested fork
gives three distinct pids (123...)
- README: document the fork mechanism, mark syscall status
|
|
|
- fix ClearKernelRAM clobbering the kernel stack (only clear $C000-$CBFF)
- fix SyscallTrap clobbering DE (buffer arg) during table dispatch
- README: run via ~/dev/gbc --headless; note bring-up bugs
|
