More Darwin.

src/libthread/Darwin.c

@@ -0,0 +1,5 @@
+int
+_schedfork(Proc *p)
+{
+	return ffork(RFMEM|RFNOWAIT, _schedinit, p);
+}

src/libthread/PowerMacintosh.c

@@ -0,0 +1,28 @@
+#include "threadimpl.h"
+
+static void
+launcherpower(int a0, int a1, int a2, int a3, int a4, int a5, int a6, int a7,
+	void (*f)(void *arg), void *arg)
+{
+	(*f)(arg);
+	threadexits(nil);
+}
+
+void
+_threadinitstack(Thread *t, void (*f)(void*), void *arg)
+{
+	ulong *tos, *stk;
+	int n;
+
+	tos = (ulong*)&t->stk[t->stksize&~7];
+	stk = tos;
+	--stk;
+	--stk;
+	--stk;
+	--stk;
+	*--stk = (ulong)arg;
+	*--stk = (ulong)f;
+	t->sched.pc = (ulong)launcherpower+LABELDPC;
+	t->sched.sp = (ulong)tos-80;
+}
+

src/libthread/asm-Darwin-PowerMacintosh.s

@@ -0,0 +1,80 @@
+/* get FPR and VR use flags with sc 0x7FF3 */
+/* get vsave with mfspr reg, 256 */
+
+.text
+.align 2
+
+.globl	__setlabel
+
+__setlabel:				/* xxx: instruction scheduling */
+	mflr	r0
+	mfcr	r5
+	mfctr	r6
+	mfxer	r7
+	stw	r0, 0*4(r3)
+	stw	r5, 1*4(r3)
+	stw	r6, 2*4(r3)
+	stw	r7, 3*4(r3)
+
+	stw	r1, 4*4(r3)
+	stw	r2, 5*4(r3)
+
+	stw	r13, (0+6)*4(r3)	/* callee-save GPRs */
+	stw	r14, (1+6)*4(r3)	/* xxx: block move */
+	stw	r15, (2+6)*4(r3)
+	stw	r16, (3+6)*4(r3)
+	stw	r17, (4+6)*4(r3)
+	stw	r18, (5+6)*4(r3)
+	stw	r19, (6+6)*4(r3)
+	stw	r20, (7+6)*4(r3)
+	stw	r21, (8+6)*4(r3)
+	stw	r22, (9+6)*4(r3)
+	stw	r23, (10+6)*4(r3)
+	stw	r24, (11+6)*4(r3)
+	stw	r25, (12+6)*4(r3)
+	stw	r26, (13+6)*4(r3)
+	stw	r27, (14+6)*4(r3)
+	stw	r28, (15+6)*4(r3)
+	stw	r29, (16+6)*4(r3)
+	stw	r30, (17+6)*4(r3)
+	stw	r31, (18+6)*4(r3)
+
+	li	r3, 0			/* return */
+	blr
+
+.globl	__gotolabel
+
+__gotolabel:
+	lwz	r13, (0+6)*4(r3)	/* callee-save GPRs */
+	lwz	r14, (1+6)*4(r3)	/* xxx: block move */
+	lwz	r15, (2+6)*4(r3)
+	lwz	r16, (3+6)*4(r3)
+	lwz	r17, (4+6)*4(r3)
+	lwz	r18, (5+6)*4(r3)
+	lwz	r19, (6+6)*4(r3)
+	lwz	r20, (7+6)*4(r3)
+	lwz	r21, (8+6)*4(r3)
+	lwz	r22, (9+6)*4(r3)
+	lwz	r23, (10+6)*4(r3)
+	lwz	r24, (11+6)*4(r3)
+	lwz	r25, (12+6)*4(r3)
+	lwz	r26, (13+6)*4(r3)
+	lwz	r27, (14+6)*4(r3)
+	lwz	r28, (15+6)*4(r3)
+	lwz	r29, (16+6)*4(r3)
+	lwz	r30, (17+6)*4(r3)
+	lwz	r31, (18+6)*4(r3)
+
+	lwz	r1, 4*4(r3)
+	lwz	r2, 5*4(r3)
+
+	lwz	r0, 0*4(r3)
+	mtlr	r0
+	lwz	r0, 1*4(r3)
+	mtcr	r0			/* mtcrf 0xFF, r0 */
+	lwz	r0, 2*4(r3)
+	mtctr	r0
+	lwz	r0, 3*4(r3)
+	mtxer	r0
+	li	r3, 1
+	blr