On 2022-04-26 and again on 2024-07-01, my PowerBook G4 failed to boot
a kernel.  Both kernels trapped early into ddb.  Both kernels worked
after a reorder, but I kept a copy of each kernel in the broken order.

In May 2022, I fixed my 1st bad kernel by modifying ofwboot to fix
the instruction cache.  This week, my 2nd bad kernel needed the same
fix (which I had never committed).  I adapted this fix from NetBSD's
ofwboot.  ok to commit?

--gkoehler

Index: arch/macppc/stand//conf.c
===================================================================
RCS file: /cvs/src/sys/arch/macppc/stand/conf.c,v
diff -u -p -r1.14 conf.c
--- arch/macppc/stand//conf.c	30 Jan 2021 14:37:01 -0000	1.14
+++ arch/macppc/stand//conf.c	6 Jul 2024 05:31:11 -0000
@@ -21,7 +21,7 @@
 #include <lib/libsa/stand.h>
 
 
-const char version[] = "1.11";
+const char version[] = "1.12";
 extern int     debug;
 
 void ofc_probe(struct consdev *);
Index: arch/macppc/stand//main.c
===================================================================
RCS file: /cvs/src/sys/arch/macppc/stand/main.c,v
diff -u -p -r1.13 main.c
--- arch/macppc/stand//main.c	23 Feb 2023 19:48:22 -0000	1.13
+++ arch/macppc/stand//main.c	6 Jul 2024 05:31:11 -0000
@@ -212,6 +212,7 @@ run_loadfile(uint64_t *marks, int howto)
 		OF_release((void*)lastpage, CLAIM_LIMIT - lastpage);
 	}
 
+	syncicache((void *)entry, (size_t)ssym - entry);
 	chain((void *)entry, bootline, ssym, esym);
 
 	_rtt();