izaya/collapseos
1
0
Fork 0
mirror of https://github.com/hsoft/collapseos.git synced 2024-12-25 05:18:06 +11:00
Code Issues Releases Wiki Activity
2481ed4b66
collapseos/emul/Makefile

112 lines
3.6 KiB
Makefile
Raw Normal View History

cfspack: make into a library Use it in shell instead of using popen()
2020-01-01 05:57:52 +11:00
CFSPACK_OBJ = ../tools/cfspack/libcfs.o
emul: add forth target
2020-03-10 13:26:02 +11:00
TARGETS = shell/shell zasm/zasm runbin/runbin forth/forth
Move "emul" folder to root
2020-01-01 05:32:09 +11:00
KERNEL = ../kernel
APPS = ../apps
zasm now bootstraps itself without scas!
2019-05-20 05:03:34 +10:00
ZASMBIN = zasm/zasm
avra: first steps
2019-12-14 09:38:40 +11:00
AVRABIN = zasm/avra
emul: add forth target
2020-03-10 13:26:02 +11:00
SHELLAPPS = zasm ed
recipes/rc2014/zasm: use BASIC shell Also made Makefile actually portable.
2019-12-12 05:01:51 +11:00
SHELLTGTS = ${SHELLAPPS:%=cfsin/%}
forth: split forth source into multiple files
2020-03-18 12:44:32 +11:00
# Those Forth source files are in a particular order
forth: remove dummy.fs All "stop pings" are now explicitly made at appropriate places. This fixes a problem I've introduced in the last commit where RAM+ was unexpectedly part of a stable ABI because of "HERE" usage in dummy.fs. This system will, anyways, soon change because boot.bin and z80c.bin will be built during the same process (but there's a bit of retooling to do before I get there).
2020-04-03 02:29:23 +11:00
FORTHSRCS = core.fs str.fs parse.fs readln.fs fmt.fs z80a.fs
Move "forth" folder to project's root folder It's not really an App.
2020-03-19 13:25:44 +11:00
FORTHSRC_PATHS = ${FORTHSRCS:%=../forth/%}
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
CFSIN_CONTENTS = $(SHELLTGTS) cfsin/user.h
OBJS = emul.o libz80/libz80.o
cfspack: make into a library Use it in shell instead of using popen()
2020-01-01 05:57:52 +11:00
SHELLOBJS = $(OBJS) $(CFSPACK_OBJ)
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
ZASMOBJS = $(SHELLOBJS)
Move apps/zasm/emul to tools/emul
2019-05-10 04:09:40 +10:00
Add tools/emul
2019-05-10 02:58:41 +10:00
.PHONY: all
avra: first steps
2019-12-14 09:38:40 +11:00
all: $(TARGETS) $(AVRABIN) $(CFSIN_CONTENTS)
Add tools/emul
2019-05-10 02:58:41 +10:00
Make the shell a userspace app That's my mega-commit you've all been waiting for. The code for the shell share more routines with userspace apps than with kernel units, because, well, its behavior is that of a userspace app, not a device driver. This created a weird situation with libraries and jump tables. Some routine belonging to the `kernel/` directory felt weird there. And then comes `apps/basic`, which will likely share even more code with the shell. I was seeing myself creating huge jump tables to reuse code from the shell. It didn't feel right. Moreover, we'll probably want basic-like apps to optionnally replace the shell. So here I am with this huge change in the project structure. I didn't test all recipes on hardware yet, I will do later. I might have broken some... But now, the structure feels better and the line between what belongs to `kernel` and what belongs to `apps` feels clearer.
2019-11-16 07:37:49 +11:00
# -o in sync with SHELL_CODE in shell/glue.asm
Replace old shell with BASIC shell fixes #80
2019-12-12 06:05:34 +11:00
shell/shell.bin: shell/glue.asm $(ZASMBIN)
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(ZASMBIN) $(KERNEL) shell/user.h $(APPS) < shell/glue.asm | tee $@ > /dev/null
Make the shell a userspace app That's my mega-commit you've all been waiting for. The code for the shell share more routines with userspace apps than with kernel units, because, well, its behavior is that of a userspace app, not a device driver. This created a weird situation with libraries and jump tables. Some routine belonging to the `kernel/` directory felt weird there. And then comes `apps/basic`, which will likely share even more code with the shell. I was seeing myself creating huge jump tables to reuse code from the shell. It didn't feel right. Moreover, we'll probably want basic-like apps to optionnally replace the shell. So here I am with this huge change in the project structure. I didn't test all recipes on hardware yet, I will do later. I might have broken some... But now, the structure feels better and the line between what belongs to `kernel` and what belongs to `apps` feels clearer.
2019-11-16 07:37:49 +11:00
Replace old shell with BASIC shell fixes #80
2019-12-12 06:05:34 +11:00
shell/shell-bin.h: shell/shell.bin
./bin2c.sh KERNEL < shell/shell.bin | tee $@ > /dev/null
basic: add fls command Also, add the new `bshell` emulated tool. BASIC is on its way to replace the shell.
2019-11-25 02:24:15 +11:00
cfspack: make into a library Use it in shell instead of using popen()
2020-01-01 05:57:52 +11:00
shell/shell: shell/shell.c $(SHELLOBJS) shell/shell-bin.h
$(CC) shell/shell.c $(SHELLOBJS) -o $@
avra: first steps
2019-12-14 09:38:40 +11:00
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
# z80c.bin is not in the prerequisites because its a bootstrap binary that
# should be updated manually through make fbootstrap.
Forth self-bootstraps!!! This commit cuts Forth's dependency on zasm. It now self-assembles its own boot-binary! Things are a bit messy now, I'll clean up later.
2020-04-02 13:19:37 +11:00
forth/forth0.bin:
cat forth/boot.bin forth/z80c.bin > $@
emul: add forth target
2020-03-10 13:26:02 +11:00
forth: add bin dict compilation stage! Big one. This allows us to write higher order words directly in Forth, which is much more convenient than writing post-immediate (see "NOT" structure in diff if you want to see what I mean) structures in ASM. These structures can then be written to ROM (rather than loaded in RAM for definitions loaded at run-time). That's quite a bit of tooling that was added, 2 compilations stages, but I think it's well worth it.
2020-03-12 15:14:44 +11:00
forth/forth0-bin.h: forth/forth0.bin
./bin2c.sh KERNEL < forth/forth0.bin | tee $@ > /dev/null
emul: add forth target
2020-03-10 13:26:02 +11:00
forth: make bin staging process a bit less hackish The goal being to add a new native code dict staging phase.
2020-03-23 12:46:43 +11:00
forth/stage1: forth/stage.c $(OBJS) forth/forth0-bin.h
$(CC) forth/stage.c $(OBJS) -o $@
forth: add bin dict compilation stage! Big one. This allows us to write higher order words directly in Forth, which is much more convenient than writing post-immediate (see "NOT" structure in diff if you want to see what I mean) structures in ASM. These structures can then be written to ROM (rather than loaded in RAM for definitions loaded at run-time). That's quite a bit of tooling that was added, 2 compilations stages, but I think it's well worth it.
2020-03-12 15:14:44 +11:00
forth: make bin staging process a bit less hackish The goal being to add a new native code dict staging phase.
2020-03-23 12:46:43 +11:00
forth/stage1dbg: forth/stage.c $(OBJS) forth/forth0-bin.h
Forth self-bootstraps!!! This commit cuts Forth's dependency on zasm. It now self-assembles its own boot-binary! Things are a bit messy now, I'll clean up later.
2020-04-02 13:19:37 +11:00
$(CC) -DDEBUG -DBOOT forth/stage.c $(OBJS) -o $@
forth: improve execution model My approach with RS was slightly wrong: RS' TOP was always containing current IP. It worked, but it was problematic when came the time to introduce RS-modifying words: it's impossible to modify RS in a word without immediately messing your flow. Therefore, what used to be RS' TOS has to be a variable that isn't changed midway by RS-modifying words. I guess that's why RS is called *return* stack...
2020-03-14 07:01:09 +11:00
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
forth/core.bin: $(FORTHSRC_PATHS) forth/stage1
forth: remove dummy.fs All "stop pings" are now explicitly made at appropriate places. This fixes a problem I've introduced in the last commit where RAM+ was unexpectedly part of a stable ABI because of "HERE" usage in dummy.fs. This system will, anyways, soon change because boot.bin and z80c.bin will be built during the same process (but there's a bit of retooling to do before I get there).
2020-04-03 02:29:23 +11:00
cat $(FORTHSRC_PATHS) ./forth/stop.fs | ./forth/stage1 | tee $@ > /dev/null
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
forth: simplify build process Make LATEST configurable through a value in the binary instead of through a ZASM constant. This enables the simple concatenation of compiled dicts and simplifies the build process.
2020-04-01 23:54:01 +11:00
forth/forth1.bin: forth/forth0.bin forth/core.bin
cat forth/forth0.bin forth/core.bin > $@
forth: add bin dict compilation stage! Big one. This allows us to write higher order words directly in Forth, which is much more convenient than writing post-immediate (see "NOT" structure in diff if you want to see what I mean) structures in ASM. These structures can then be written to ROM (rather than loaded in RAM for definitions loaded at run-time). That's quite a bit of tooling that was added, 2 compilations stages, but I think it's well worth it.
2020-03-12 15:14:44 +11:00
forth/forth1-bin.h: forth/forth1.bin
./bin2c.sh KERNEL < forth/forth1.bin | tee $@ > /dev/null
forth: Forth-ify ROT, a native word! This requires us to significantly adjust our build process, which now has 3 stages.
2020-03-25 04:46:05 +11:00
forth/stage2: forth/stage.c $(OBJS) forth/forth1-bin.h
$(CC) -DSTAGE2 forth/stage.c $(OBJS) -o $@
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
forth/forth: forth/forth.c $(OBJS) forth/forth1-bin.h
emul: add forth target
2020-03-10 13:26:02 +11:00
$(CC) forth/forth.c $(OBJS) -o $@
tools/emul: re-organize .h files What used to be `tools/emul/user.h` was in fact specific to zasm, so I moved it there. To avoid name confusion, I renamed what used to be kernel.h and user.h to kernel-bin.h and user-bin.h.
2019-06-03 22:12:44 +10:00
zasm/kernel-bin.h: zasm/kernel.bin
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
./bin2c.sh KERNEL < zasm/kernel.bin | tee $@ > /dev/null
zasm now bootstraps itself without scas!
2019-05-20 05:03:34 +10:00
tools/emul: re-organize .h files What used to be `tools/emul/user.h` was in fact specific to zasm, so I moved it there. To avoid name confusion, I renamed what used to be kernel.h and user.h to kernel-bin.h and user-bin.h.
2019-06-03 22:12:44 +10:00
zasm/zasm-bin.h: zasm/zasm.bin
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
./bin2c.sh USERSPACE < zasm/zasm.bin | tee $@ > /dev/null
Add tools/emul
2019-05-10 02:58:41 +10:00
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(ZASMBIN): zasm/zasm.c $(ZASMOBJS) zasm/kernel-bin.h zasm/zasm-bin.h
$(CC) zasm/zasm.c $(ZASMOBJS) -o $@
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
avra: first steps
2019-12-14 09:38:40 +11:00
zasm/avra.bin: $(ZASMBIN)
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(ZASMBIN) $(KERNEL) $(APPS) zasm/user.h < $(APPS)/zasm/gluea.asm > $@
avra: first steps
2019-12-14 09:38:40 +11:00
zasm/avra-bin.h: zasm/avra.bin
./bin2c.sh USERSPACE < zasm/avra.bin | tee $@ > /dev/null
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(AVRABIN): zasm/zasm.c $(ZASMOBJS) zasm/kernel-bin.h zasm/avra-bin.h
$(CC) -D AVRA zasm/zasm.c $(ZASMOBJS) -o $@
avra: first steps
2019-12-14 09:38:40 +11:00
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
runbin/runbin: runbin/runbin.c $(OBJS)
$(CC) runbin/runbin.c $(OBJS) -o $@
Add tools/emul
2019-05-10 02:58:41 +10:00
libz80/libz80.o: libz80/z80.c
Make tools/emul *BSD-friendly
2019-07-12 11:21:54 +10:00
$(MAKE) -C libz80/codegen opcodes
zasm: add support for .fill directive
2019-05-20 22:20:00 +10:00
$(CC) -Wall -ansi -g -c -o libz80/libz80.o libz80/z80.c
Add tools/emul
2019-05-10 02:58:41 +10:00
tools/emul: deduplicate a little bit of C code
2019-12-03 08:35:49 +11:00
emul.o: emul.c
$(CC) -c -o emul.o emul.c
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(CFSPACK_OBJ): ${@:%.o=%.c}
Move "emul" folder to root
2020-01-01 05:32:09 +11:00
$(MAKE) -C ../tools/cfspack
tools/emul/shell: add filesystem support With the help of the newly-introduced cfspack tool, we can mount a filesystem in our emulated shell and play around. Read-only for now. Unpacking incoming.
2019-05-13 01:20:31 +10:00
Make the shell a userspace app That's my mega-commit you've all been waiting for. The code for the shell share more routines with userspace apps than with kernel units, because, well, its behavior is that of a userspace app, not a device driver. This created a weird situation with libraries and jump tables. Some routine belonging to the `kernel/` directory felt weird there. And then comes `apps/basic`, which will likely share even more code with the shell. I was seeing myself creating huge jump tables to reuse code from the shell. It didn't feel right. Moreover, we'll probably want basic-like apps to optionnally replace the shell. So here I am with this huge change in the project structure. I didn't test all recipes on hardware yet, I will do later. I might have broken some... But now, the structure feels better and the line between what belongs to `kernel` and what belongs to `apps` feels clearer.
2019-11-16 07:37:49 +11:00
# -o in sync with USER_CODE in shell/user.h
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
$(SHELLTGTS): $(ZASMBIN)
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(ZASMBIN) -o 42 $(KERNEL) $(APPS) shell/user.h < $(APPS)/${@:cfsin/%=%}/glue.asm > $@
tools/emul: make shell's cfsin always have all apps compiled in it
2019-06-03 22:32:25 +10:00
tools/emul/shell: copy user.h in its cfsin This way, we can compile apps from within the shell that run on the shell.
2019-06-03 23:25:47 +10:00
cfsin/user.h: shell/user.h
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
cp shell/user.h $@
tools/emul/shell: copy user.h in its cfsin This way, we can compile apps from within the shell that run on the shell.
2019-06-03 23:25:47 +10:00
zasm now bootstraps itself without scas!
2019-05-20 05:03:34 +10:00
.PHONY: updatebootstrap
avra: first steps
2019-12-14 09:38:40 +11:00
updatebootstrap: $(ZASMBIN)
emul/zasm: use libcfs This allows us to get rid of the zasm.sh wrapper.
2020-01-01 07:07:39 +11:00
$(ZASMBIN) $(KERNEL) < zasm/glue.asm > zasm/kernel.bin
$(ZASMBIN) $(KERNEL) $(APPS) zasm/user.h < $(APPS)/zasm/glue.asm > zasm/zasm.bin
zasm now bootstraps itself without scas!
2019-05-20 05:03:34 +10:00
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
# We need to double wrap around dummy.fs because at stage3, we have high-level
# words and they write to HERE at initialization.
forth: Forth-ify ROT, a native word! This requires us to significantly adjust our build process, which now has 3 stages.
2020-03-25 04:46:05 +11:00
.PHONY: fbootstrap
forth: improve bootstrap process
2020-03-25 12:13:02 +11:00
fbootstrap: forth/stage2
Forth self-bootstraps!!! This commit cuts Forth's dependency on zasm. It now self-assembles its own boot-binary! Things are a bit messy now, I'll clean up later.
2020-04-02 13:19:37 +11:00
cat ./forth/conf.fs ../forth/boot.fs | ./forth/stage2 > forth/boot.bin
forth: remove dummy.fs All "stop pings" are now explicitly made at appropriate places. This fixes a problem I've introduced in the last commit where RAM+ was unexpectedly part of a stable ABI because of "HERE" usage in dummy.fs. This system will, anyways, soon change because boot.bin and z80c.bin will be built during the same process (but there's a bit of retooling to do before I get there).
2020-04-03 02:29:23 +11:00
cat ../forth/z80c.fs forth/emul.fs ../forth/icore.fs | ./forth/stage2 | tee forth/z80c.bin > /dev/null
forth: Forth-ify ROT, a native word! This requires us to significantly adjust our build process, which now has 3 stages.
2020-03-25 04:46:05 +11:00
Add tools/emul
2019-05-10 02:58:41 +10:00
.PHONY: clean
clean:
Make makefiles and shell scripts portable It's no longer required to use `gmake` under OpenBSD and shell scripts no longer require bash.
2019-12-10 01:36:38 +11:00
rm -f $(TARGETS) $(SHELLTGTS) emul.o zasm/*-bin.h shell/*-bin.h
Reference in New Issue Copy Permalink