restore large xfer workaround

I'm sick of the way the simple extension builder in setup.py doesn't
understand header dependencies.

CMakeLists.txt

@@ -0,0 +1,12 @@
+cmake_minimum_required(VERSION 3.4...3.18)
+project("_piomatter")
+set(CMAKE_INTERPROCEDURAL_OPTIMIZATION FALSE)
+
+find_package(pybind11 CONFIG)
+pybind11_add_module("_piomatter"
+    src/pymain.cpp src/piolib/pio_rp1.c src/piolib/piolib.c)
+target_include_directories("_piomatter" PRIVATE
+    ${CMAKE_CURRENT_LIST_DIR}/src/include
+    ${CMAKE_CURRENT_LIST_DIR}/src/piolib/include)
+set_property(TARGET "_piomatter" PROPERTY
+    CXX_STANDARD 20)

pyproject.toml

@@ -3,6 +3,7 @@ requires = [
     "setuptools>=42",
     "pybind11>=2.10.0",
     "setuptools_scm[toml]>=6.2",
+    "cmake>=3.12"
 ]
 build-backend = "setuptools.build_meta"
 

setup.py

@@ -1,36 +1,80 @@
 # Available at setup time due to pyproject.toml
-from pybind11.setup_helpers import Pybind11Extension, build_ext
-from setuptools import setup
+import multiprocessing
+import os
+import subprocess
+import sys
+from pathlib import Path
+
+from setuptools import Extension, setup
+from setuptools.command.build_ext import build_ext
 from setuptools_scm import get_version
 
 __version__ = get_version()
 
+# A CMakeExtension needs a sourcedir instead of a file list.
+# The name must be the _single_ output extension from the CMake build.
+# If you need multiple extensions, see scikit-build.
+class CMakeExtension(Extension):
+    def __init__(self, name: str, sourcedir: str = "") -> None:
+        super().__init__(name, sources=[])
+        self.sourcedir = os.fspath(Path(sourcedir).resolve())
+
+
+class CMakeBuild(build_ext):
+    def build_extension(self, ext: CMakeExtension) -> None:
+        # Must be in this form due to bug in .resolve() only fixed in Python 3.10+
+        ext_fullpath = Path.cwd() / self.get_ext_fullpath(ext.name)
+        extdir = ext_fullpath.parent.resolve()
+
+        # Using this requires trailing slash for auto-detection & inclusion of
+        # auxiliary "native" libs
+
+        debug = int(os.environ.get("DEBUG", 0)) if self.debug is None else self.debug
+        cfg = "Debug" if debug else "Release"
+
+        # Set Python_EXECUTABLE instead if you use PYBIND11_FINDPYTHON
+        # EXAMPLE_VERSION_INFO shows you how to pass a value into the C++ code
+        # from Python.
+        cmake_args = [
+            "-GUnix Makefiles",
+            f"-DCMAKE_LIBRARY_OUTPUT_DIRECTORY={extdir}{os.sep}",
+            f"-DPYTHON_EXECUTABLE={sys.executable}",
+            f"-DCMAKE_BUILD_TYPE={cfg}",  # not used on MSVC, but no harm
+        ]
+        build_args = []
+        # Adding CMake arguments set as environment variable
+        # (needed e.g. to build for ARM OSx on conda-forge)
+        if "CMAKE_ARGS" in os.environ:
+            cmake_args += [item for item in os.environ["CMAKE_ARGS"].split(" ") if item]
+
+        build_args += [f"-j{multiprocessing.cpu_count()}"]
+
+        build_temp = Path(self.build_temp) / ext.name
+        if not build_temp.exists():
+            build_temp.mkdir(parents=True)
+
+        subprocess.run(
+            ["cmake", ext.sourcedir, *cmake_args], cwd=build_temp, check=True
+        )
+        subprocess.run(
+            ["make", *build_args], cwd=build_temp, check=True
+        )
+
+
+
 # The main interface is through Pybind11Extension.
 # * You can add cxx_std=11/14/17, and then build_ext can be removed.
 # * You can set include_pybind11=false to add the include directory yourself,
 #   say from a submodule.
 
-ext_modules = [
-    Pybind11Extension("adafruit_blinka_raspberry_pi5_piomatter._piomatter",
-        ["src/pymain.cpp", "src/piolib/piolib.c", "src/piolib/pio_rp1.c"],
-        define_macros = [('VERSION_INFO', __version__)],
-        include_dirs = ['./src/include', './src/piolib/include'],
-        cxx_std=20,
-        # use this setting when debugging
-        extra_compile_args = ["-g3", "-Og"],
-        ),
-]
-
 setup(
     name="Adafruit-Blinka-Raspberry-Pi5-Piomatter",
     version=__version__,
     url="https://github.com/adafruit/Adafruit_Blinka_Raspberry_Pi5_Piomatter",
     description="HUB75 matrix driver for Raspberry Pi 5 using PIO",
     long_description="A pio-based driver",
-    ext_modules=ext_modules,
-    # Currently, build_ext only provides an optional "highest supported C++
-    # level" feature, but in the future it may provide more features.
-    cmdclass={"build_ext": build_ext},
+    ext_modules=[CMakeExtension("adafruit_blinka_raspberry_pi5_piomatter._piomatter")],
+    cmdclass={"build_ext": CMakeBuild},
     zip_safe=False,
     python_requires=">=3.11",
     packages=['adafruit_blinka_raspberry_pi5_piomatter'],

src/include/piomatter/piomatter.h

@@ -12,6 +12,34 @@
 
 namespace piomatter {
 
+static int pio_sm_xfer_data_large(PIO pio, int sm, int direction, size_t size,
+                                  uint32_t *databuf) {
+    int r = pio_sm_xfer_data(pio, sm, direction, size, databuf);
+    if (r != -EPERM) {
+        // Kernels that don't support large transfers return EPERM
+        return r;
+    }
+    static bool once = false;
+    if (!once) {
+        once = true;
+        fprintf(stderr,
+                "Transmission limit workaround engaged. May reduce quality of "
+                "output.\nSee https://github.com/raspberrypi/utils/issues/123 "
+                "for details.");
+    }
+    constexpr size_t MAX_XFER = 65532;
+    while (size) {
+        size_t xfersize = std::min(size_t{MAX_XFER}, size);
+        r = pio_sm_xfer_data(pio, sm, direction, xfersize, databuf);
+        if (r) {
+            return r;
+        }
+        size -= xfersize;
+        databuf += xfersize / sizeof(*databuf);
+    }
+    return 0;
+}
+
 static uint64_t monotonicns64() {
     struct timespec tp;
     clock_gettime(CLOCK_MONOTONIC, &tp);
@@ -181,8 +209,8 @@ struct piomatter : piomatter_base {
                 auto dataptr = const_cast<uint32_t *>(&data[0]);
                 // returns err = rp1_ioctl.... which seems to be a negative
                 // errno value
-                int r =
-                    pio_sm_xfer_data(pio, sm, PIO_DIR_TO_SM, datasize, dataptr);
+                int r = pio_sm_xfer_data_large(pio, sm, PIO_DIR_TO_SM, datasize,
+                                               dataptr);
                 if (r != 0) {
                     pending_error_errno.store(-r);
                     printf("xfer_data() returned error %d (%s)\n", r,