Remove the notion of "sys*"

- This idea originates from very early in the project and was, at the
  time, a very easy way to categorise things.
- Now, it doesn't really make much sense - it is fairly arbitary, often
  occuring when there is a change in kernel, but not from builder-hex0
  to fiwix, and sysb is in reality completely unnecessary.
- In short, the sys* stuff is a bit of a mess that makes the project
  more difficult to understand.
- This puts everything down into one folder and has a manifest file that
  is used to generate the build scripts on the fly rather than using
  coded scripts.
- This is created in the "seed" stage.

stage0-posix -- (calls) --> seed -- (generates) --> main steps

Alongside this change there are a variety of other smaller fixups to the
general structure of the live-bootstrap rootfs.

- Creating a rootfs has become much simpler and is defined as code in
  go.sh. The new structure, for an about-to-be booted system, is

/
-- /steps (direct copy of steps/)
-- /distfiles (direct copy of distfiles/)
-- all files from seed/*
-- all files from seed/stage0-posix/*

- There is no longer such a thing as /usr/include/musl, this didn't
  really make any sense, as musl is the final libc used. Rather, to
  separate musl and mes, we have /usr/include/mes, which is much easier
  to work with.
- This also makes mes easier to blow away later.
- A few things that weren't properly in packages have been changed;
  checksum-transcriber, simple-patch, kexec-fiwix have all been given
  fully qualified package names.
- Highly breaking change, scripts now exist in their package directory
  but NOT WITH THE packagename.sh. Rather, they use pass1.sh, pass2.sh,
  etc. This avoids manual definition of passes.
  - Ditto with patches; default directory is patches, but then any patch
    series specific to a pass are named patches-passX.
This commit is contained in:
fosslinux 2023-11-07 10:51:23 +11:00
parent 0907cfd073
commit 6ed2e09f3a
546 changed files with 700 additions and 1299 deletions

View file

@ -0,0 +1,28 @@
SPDX-FileCopyrightText: 2022 fosslinux <fosslinux@aussies.space>
SPDX-License-Identifier: PSF-2.0
frozenset() is a feature only added in Python 2.5, but we are
building Python 2.5, so we must bypass this logic. (It is not
critical, so we can just remove it).
--- Lib/keyword.py 2022-10-11 12:51:13.050744758 +1100
+++ Lib/keyword.py 2022-10-11 12:52:05.946372559 +1100
@@ -10,7 +10,7 @@
python Lib/keyword.py
"""
-__all__ = ["iskeyword", "kwlist"]
+__all__ = ["kwlist"]
kwlist = [
#--start keywords--
@@ -48,8 +48,6 @@
#--end keywords--
]
-iskeyword = frozenset(kwlist).__contains__
-
def main():
import sys, re

View file

@ -0,0 +1,36 @@
SPDX-FileCopyrightText: 2011 Antoine Pitrou <solipsis@pitrou.net>
SPDX-License-Identifier: PSF-2.0
Don't run pgen twice when using make -j
Don't ignore pgen error (on "make Parser/pgen.stamp")
diff --git a/Makefile.pre.in b/Makefile.pre.in
index 25a82aff01..03f85fd04b 100644
--- Makefile.pre.in
+++ Makefile.pre.in
@@ -471,10 +471,12 @@ Modules/getpath.o: $(srcdir)/Modules/getpath.c Makefile
Modules/python.o: $(srcdir)/Modules/python.c
$(MAINCC) -c $(PY_CFLAGS) -o $@ $(srcdir)/Modules/python.c
-
-$(GRAMMAR_H) $(GRAMMAR_C): $(PGEN) $(GRAMMAR_INPUT)
+# Use a stamp file to prevent make -j invoking pgen twice
+$(GRAMMAR_H) $(GRAMMAR_C): Parser/pgen.stamp
+Parser/pgen.stamp: $(PGEN) $(GRAMMAR_INPUT)
-@ mkdir Include
- -$(PGEN) $(GRAMMAR_INPUT) $(GRAMMAR_H) $(GRAMMAR_C)
+ $(PGEN) $(GRAMMAR_INPUT) $(GRAMMAR_H) $(GRAMMAR_C)
+ -touch Parser/pgen.stamp
$(PGEN): $(PGENOBJS)
$(CC) $(OPT) $(LDFLAGS) $(PGENOBJS) $(LIBS) -o $(PGEN)
@@ -1030,7 +1032,7 @@ clean: pycremoval
clobber: clean
-rm -f $(BUILDPYTHON) $(PGEN) $(LIBRARY) $(LDLIBRARY) $(DLLLIBRARY) \
- tags TAGS \
+ tags TAGS Parser/pgen.stamp \
config.cache config.log pyconfig.h Modules/config.c
-rm -rf build platform
-rm -rf $(PYTHONFRAMEWORKDIR)

View file

@ -0,0 +1,33 @@
SPDX-FileCopyrightText: 2022 fosslinux <fosslinux@aussies.space>
SPDX-License-Identifier: PSF-2.0
musl (correctly) implements the POSIX posix_close function, however
this was added after Python 2.5.6 was released.
--- Modules/posixmodule.c 2022-10-05 18:38:46.718131893 +1100
+++ Modules/posixmodule.c 2022-10-05 18:39:07.049250312 +1100
@@ -5208,12 +5208,12 @@
}
-PyDoc_STRVAR(posix_close__doc__,
+PyDoc_STRVAR(py_posix_close__doc__,
"close(fd)\n\n\
Close a file descriptor (for low level IO).");
static PyObject *
-posix_close(PyObject *self, PyObject *args)
+py_posix_close(PyObject *self, PyObject *args)
{
int fd, res;
if (!PyArg_ParseTuple(args, "i:close", &fd))
@@ -7371,7 +7371,7 @@
{"tcsetpgrp", posix_tcsetpgrp, METH_VARARGS, posix_tcsetpgrp__doc__},
#endif /* HAVE_TCSETPGRP */
{"open", posix_open, METH_VARARGS, posix_open__doc__},
- {"close", posix_close, METH_VARARGS, posix_close__doc__},
+ {"close", py_posix_close, METH_VARARGS, py_posix_close__doc__},
{"dup", posix_dup, METH_VARARGS, posix_dup__doc__},
{"dup2", posix_dup2, METH_VARARGS, posix_dup2__doc__},
{"lseek", posix_lseek, METH_VARARGS, posix_lseek__doc__},

View file

@ -0,0 +1,29 @@
SPDX-FileCopyrightText: 2022 fosslinux <fosslinux@aussies.space>
SPDX-License-Identifier: PSF-2.0
sorted() was only added in Python 2.5. But we are building Python 2.5.
We cannot use .sort(), as it doesn't support the key= parameter.
Instead we just use a basic custom selection sort to sort it ourselves
using a custom key.
--- Tools/compiler/astgen.py.bak 2022-07-11 09:24:59.600238862 +1000
+++ Tools/compiler/astgen.py 2022-07-11 09:32:25.814974174 +1000
@@ -215,7 +215,15 @@
# some extra code for a Node's __init__ method
name = mo.group(1)
cur = classes[name]
- return sorted(classes.values(), key=lambda n: n.name)
+ ret = classes.values()
+ # basic custom selection sort
+ for i in range(len(ret)):
+ min_i = i
+ for j in range(i + 1, len(ret)):
+ if ret[min_i].name > ret[j].name:
+ min_i = j
+ ret[i], ret[min_i] = ret[min_i], ret[i]
+ return ret
def main():
prologue, epilogue = load_boilerplate(sys.argv[-1])

View file

@ -0,0 +1,20 @@
SPDX-FileCopyrightText: 2022 fosslinux <fosslinux@aussies.space>
SPDX-License-Identifier: PSF-2.0
Again, Python 2.5 added the key= argument for sorting functions,
which is not available when we are building Python 2.5.
Sorting is absolutely unnessecary when generating defines for a
header file so we can just remove it.
--- Lib/sre_constants.py 2004-08-25 12:22:30.000000000 +1000
+++ Lib/sre_constants.py 2022-10-09 20:18:40.332233858 +1100
@@ -219,7 +219,6 @@
if __name__ == "__main__":
def dump(f, d, prefix):
items = d.items()
- items.sort(key=lambda a: a[1])
for k, v in items:
f.write("#define %s_%s %s\n" % (prefix, k.upper(), v))
f = open("sre_constants.h", "w")