libtf: initial commit

libtf is to be user-space cooperative threading library similar
to State Threads (http://state-threads.sourceforge.net/), but
with additional support for multiple cores, using better
algorithms and taking advantage of new Linux kernel syscalls
such as eventfd, signalfd and epoll (edge-triggered mode).

Initial implementation has setjmp based user-space context
switching and trivial testcase. Works on Linux/x86.

TFbuild uses ideas from different build systems, namely Kbuild,
but it's inner workings are quite different. All build files are
included (using macro trickery) instead of recursive making. Thus
the build dependency graph is complete and should yield good make
performance. Also parallel stuff should work.

2 files changed, 35 insertions, 0 deletions

diff --git a/test/TFbuild b/test/TFbuild
new file mode 100644
index 0000000..d2648ed
--- /dev/null
+++ b/test/TFbuild
@@ -0,0 +1,3 @@
+progs-$(TEST)		+= simple1 sleep
+
+LIBS			+= $(objtree)src/libtf.a
diff --git a/test/simple1.c b/test/simple1.c
new file mode 100644
index 0000000..65a787c
--- /dev/null
+++ b/test/simple1.c
@@ -0,0 +1,32 @@
+#include <libtf/tf.h>
+#include <stdio.h>
+
+struct ctx {
+	int id;
+};
+
+static void work_fiber(void *ptr)
+{
+	struct ctx *c = (struct ctx*) ptr;
+
+	printf("Hello%d.1\n", c->id);
+	tf_yield();
+	printf("Hello%d.2\n", c->id);
+}
+
+static void init_fiber(void *ptr)
+{
+	struct ctx *c;
+	int i;
+
+	for (i = 0; i < 6; i++) {
+		c = tf_fiber_create(work_fiber, sizeof(struct ctx));
+		c->id = i;
+		tf_fiber_put(c);
+	}
+}
+
+int main(int argc, char **argv)
+{
+	return tf_main(init_fiber);
+}