Bring in two bandaids to get the elf trampoline to work again, until I find

a proper solution.
- Add a dummy entry point which just calls the C entry points, and try to make
sure it's the first code in the binary.
- Copy a bit more than func_end to try to copy the whole load_kernel()
function. gcc4 puts code behind the func_end symbol.

Approved by:	re (blanket)

sys/arm/arm/elf_trampoline.c

@@ -48,6 +48,7 @@ extern char kernel_end[];
 extern void *_end;
 
 void __start(void);
+void __startC(void);
 
 #define GZ_HEAD	0xa
 
@@ -66,6 +67,13 @@ void __start(void);
   defined(CPU_XSCALE_PXA2X0) || defined(CPU_XSCALE_IXP425) ||	\
   defined(CPU_XSCALE_80219)
 #define cpu_idcache_wbinv_all	xscale_cache_purgeID
+#elif defined(CPU_XSCALE_81342)
+#define cpu_idcache_wbinv_all	xscalec3_cache_purgeID
+#endif
+#ifdef CPU_XSCALE_81342
+#define cpu_l2cache_wbinv_all	xscalec3_l2cache_purge
+#else
+#define cpu_l2cache_wbinv_all()	
 #endif
 
 
@@ -138,7 +146,7 @@ bzero(void *addr, int count)
 static void arm9_setup(void);
 
 void
-_start(void)
+_startC(void)
 {
 	int physaddr = KERNPHYSADDR;
 	int tmp1;
@@ -207,6 +215,7 @@ _start(void)
 		arm9_setup();
 #endif
 	cpu_idcache_wbinv_all();
+	cpu_l2cache_wbinv_all();
 #endif
 	__start();
 }
@@ -520,7 +529,7 @@ load_kernel(unsigned int kstart, unsigned int curaddr,unsigned int func_end,
 extern char func_end[];
 
 
-#define PMAP_DOMAIN_KERNEL	15 /*
+#define PMAP_DOMAIN_KERNEL	0 /*
 				    * Just define it instead of including the
 				    * whole VM headers set.
 				    */
@@ -595,10 +604,11 @@ __start(void)
 		kernel = (char *)&_end;
 		altdst = 4 + load_kernel((unsigned int)kernel, 
 		    (unsigned int)curaddr,
-		    (unsigned int)&func_end , 0);
+		    (unsigned int)&func_end + 800 , 0);
 		if (altdst > dst)
 			dst = altdst;
 		cpu_idcache_wbinv_all();
+		cpu_l2cache_wbinv_all();
 		__asm __volatile("mrc p15, 0, %0, c1, c0, 0\n"
 		    "bic %0, %0, #1\n" /* MMU_ENABLE */
 		    "mcr p15, 0, %0, c1, c0, 0\n"
@@ -616,7 +626,7 @@ __start(void)
 	sp = sp &~3;
 	dst = (void *)(sp + 4);
 	memcpy((void *)dst, (void *)&load_kernel, (unsigned int)&func_end - 
-	    (unsigned int)&load_kernel);
+	    (unsigned int)&load_kernel + 800);
 	do_call(dst, kernel, dst + (unsigned int)(&func_end) - 
-	    (unsigned int)(&load_kernel), sp);
+	    (unsigned int)(&load_kernel) + 800, sp);
 }

sys/conf/Makefile.arm

@@ -66,23 +66,27 @@ SYSTEM_LD_TAIL +=;sed s/" + SIZEOF_HEADERS"// ldscript.$M\
 FILES_CPU_FUNC =	$S/$M/$M/cpufunc_asm_arm7tdmi.S \
 	$S/$M/$M/cpufunc_asm_arm8.S $S/$M/$M/cpufunc_asm_arm9.S \
 	$S/$M/$M/cpufunc_asm_sa1.S $S/$M/$M/cpufunc_asm_arm10.S \
-	$S/$M/$M/cpufunc_asm_xscale.S $S/$M/$M/cpufunc_asm.S
+	$S/$M/$M/cpufunc_asm_xscale.S $S/$M/$M/cpufunc_asm.S \
+	$S/$M/$M/cpufunc_asm_xscale_c3.S
 trampoline: ${KERNEL_KO}.tramp
 ${KERNEL_KO}.tramp: ${KERNEL_KO} $S/$M/$M/inckern.S $S/$M/$M/elf_trampoline.c
 	echo "#define KERNNAME \"${KERNEL_KO}.tmp\"" >opt_kernname.h 
 	sed s/${KERNVIRTADDR}/${KERNPHYSADDR}/ ldscript.$M > ldscript.$M.tramp
 	sed s/" + SIZEOF_HEADERS"// ldscript.$M.tramp > \
 	ldscript.$M.tramp.noheader
+	echo "#include <machine/asm.h>" >tmphack.S
+	echo "ENTRY(_start)" >>tmphack.S
+	echo "bl _startC" >>tmphack.S
 	${OBJCOPY} --strip-symbol '$$d' --strip-symbol '$$a' \
 	-g --strip-symbol '$$t' ${FULLKERNEL} ${KERNEL_KO}.tmp
 	eval $$(stat -s ${KERNEL_KO}.tmp) && \
 	echo "#define KERNSIZE $$st_size" >>opt_kernname.h
 	${CC} -O -nostdlib -I. -Xlinker -T -Xlinker ldscript.$M.tramp \
-	$S/$M/$M/elf_trampoline.c $S/$M/$M/inckern.S ${FILES_CPU_FUNC} \
+	tmphack.S $S/$M/$M/elf_trampoline.c $S/$M/$M/inckern.S \
-	-o ${KERNEL_KO}.tramp
+	${FILES_CPU_FUNC} -o ${KERNEL_KO}.tramp
 	${CC} -O -nostdlib -I. -Xlinker -T -Xlinker ldscript.$M.tramp.noheader \
-	$S/$M/$M/elf_trampoline.c $S/$M/$M/inckern.S ${FILES_CPU_FUNC} -o \
+	tmphack.S $S/$M/$M/elf_trampoline.c $S/$M/$M/inckern.S \
-	${KERNEL_KO}.tramp.noheader 
+	${FILES_CPU_FUNC} -o ${KERNEL_KO}.tramp.noheader 
 	${OBJCOPY} -S -O binary ${KERNEL_KO}.tramp.noheader \
 	${KERNEL_KO}.tramp.bin 
 	${OBJCOPY} ${STRIP_FLAGS} ${KERNEL_KO}.tmp 
@@ -95,15 +99,15 @@ ${KERNEL_KO}.tramp: ${KERNEL_KO} $S/$M/$M/inckern.S $S/$M/$M/elf_trampoline.c
 	echo "#define KERNCOMPSIZE $$st_size" >>opt_kernname.h
 	${CC} -O2 -DKZIP -I. -c $S/kern/inflate.c -o inflate-tramp.o
 	${CC} -O -nostdlib -I. -Xlinker -T -Xlinker ldscript.$M.tramp \
-	-DKZIP $S/$M/$M/elf_trampoline.c inflate-tramp.o $S/$M/$M/inckern.S \
+	-DKZIP tmphack.S $S/$M/$M/elf_trampoline.c inflate-tramp.o \
-	${FILES_CPU_FUNC} -o ${KERNEL_KO}.gz.tramp
+	$S/$M/$M/inckern.S ${FILES_CPU_FUNC} -o ${KERNEL_KO}.gz.tramp
 	${CC} -O -nostdlib -I. -Xlinker -T -Xlinker ldscript.$M.tramp.noheader \
-	-DKZIP $S/$M/$M/elf_trampoline.c inflate-tramp.o $S/$M/$M/inckern.S \
+	-DKZIP tmphack.S $S/$M/$M/elf_trampoline.c inflate-tramp.o \
-	${FILES_CPU_FUNC} -o ${KERNEL_KO}.tramp.noheader
+	$S/$M/$M/inckern.S ${FILES_CPU_FUNC} -o ${KERNEL_KO}.tramp.noheader
 	${OBJCOPY} -S -O binary ${KERNEL_KO}.tramp.noheader \
 	${KERNEL_KO}.gz.tramp.bin
 	rm ${KERNEL_KO}.tmp.gz ${KERNEL_KO}.tramp.noheader opt_kernname.h \
-	inflate-tramp.o
+	inflate-tramp.o tmphack.S
 
 %BEFORE_DEPEND