Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6

Conflicts:

	arch/sparc/kernel/of_device.c
diff --git a/arch/sparc/Makefile b/arch/sparc/Makefile
index 6668e60..9592889 100644
--- a/arch/sparc/Makefile
+++ b/arch/sparc/Makefile
@@ -52,20 +52,24 @@
 export INIT_Y CORE_Y DRIVERS_Y NET_Y LIBS_Y HEAD_Y kallsyms.o
 
 # Default target
-all: image
+all: zImage
 
 boot := arch/sparc/boot
 
-image tftpboot.img: vmlinux
+image zImage tftpboot.img: vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $(boot)/$@
 
 archclean:
 	$(Q)$(MAKE) $(clean)=$(boot)
 
+# This is the image used for packaging
+KBUILD_IMAGE := $(boot)/zImage
+
 CLEAN_FILES += arch/$(ARCH)/boot/System.map
 
 # Don't use tabs in echo arguments.
 define archhelp
   echo  '* image        - kernel image ($(boot)/image)'
+  echo  '* zImage       - stripped kernel image ($(boot)/zImage)'
   echo  '  tftpboot.img - image prepared for tftp'
 endef
diff --git a/arch/sparc/boot/Makefile b/arch/sparc/boot/Makefile
index 22d331e..3e77a9f 100644
--- a/arch/sparc/boot/Makefile
+++ b/arch/sparc/boot/Makefile
@@ -19,6 +19,9 @@
       cmd_sysmap        = $(CONFIG_SHELL) $(srctree)/scripts/mksysmap
 quiet_cmd_image = LD      $@
       cmd_image = $(LD) $(LDFLAGS) $(EXTRA_LDFLAGS) $(LDFLAGS_$(@F)) -o $@
+quiet_cmd_strip = STRIP   $@
+      cmd_strip = $(STRIP) -R .comment -R .note -K sun4u_init -K _end -K _start $(obj)/image -o $@
+
 
 define rule_image
 	$(if $($(quiet)cmd_image),               \
@@ -49,6 +52,9 @@
 $(obj)/image: $(obj)/btfix.o FORCE
 	$(call if_changed_rule,image)
 
+$(obj)/zImage: $(obj)/image
+	$(call if_changed,strip)
+
 $(obj)/tftpboot.img: $(obj)/piggyback $(obj)/System.map $(obj)/image FORCE
 	$(call if_changed,elftoaout)
 	$(call if_changed,piggy)
diff --git a/arch/sparc/include/asm/smp_32.h b/arch/sparc/include/asm/smp_32.h
index 7201752..a8180e5 100644
--- a/arch/sparc/include/asm/smp_32.h
+++ b/arch/sparc/include/asm/smp_32.h
@@ -50,27 +50,24 @@
 void smp_bogo(struct seq_file *);
 void smp_info(struct seq_file *);
 
-BTFIXUPDEF_CALL(void, smp_cross_call, smpfunc_t, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long)
+BTFIXUPDEF_CALL(void, smp_cross_call, smpfunc_t, cpumask_t, unsigned long, unsigned long, unsigned long, unsigned long)
 BTFIXUPDEF_CALL(int, __hard_smp_processor_id, void)
 BTFIXUPDEF_BLACKBOX(hard_smp_processor_id)
 BTFIXUPDEF_BLACKBOX(load_current)
 
-#define smp_cross_call(func,arg1,arg2,arg3,arg4,arg5) BTFIXUP_CALL(smp_cross_call)(func,arg1,arg2,arg3,arg4,arg5)
+#define smp_cross_call(func,mask,arg1,arg2,arg3,arg4) BTFIXUP_CALL(smp_cross_call)(func,mask,arg1,arg2,arg3,arg4)
 
-static inline void xc0(smpfunc_t func) { smp_cross_call(func, 0, 0, 0, 0, 0); }
+static inline void xc0(smpfunc_t func) { smp_cross_call(func, cpu_online_map, 0, 0, 0, 0); }
 static inline void xc1(smpfunc_t func, unsigned long arg1)
-{ smp_cross_call(func, arg1, 0, 0, 0, 0); }
+{ smp_cross_call(func, cpu_online_map, arg1, 0, 0, 0); }
 static inline void xc2(smpfunc_t func, unsigned long arg1, unsigned long arg2)
-{ smp_cross_call(func, arg1, arg2, 0, 0, 0); }
+{ smp_cross_call(func, cpu_online_map, arg1, arg2, 0, 0); }
 static inline void xc3(smpfunc_t func, unsigned long arg1, unsigned long arg2,
 			   unsigned long arg3)
-{ smp_cross_call(func, arg1, arg2, arg3, 0, 0); }
+{ smp_cross_call(func, cpu_online_map, arg1, arg2, arg3, 0); }
 static inline void xc4(smpfunc_t func, unsigned long arg1, unsigned long arg2,
 			   unsigned long arg3, unsigned long arg4)
-{ smp_cross_call(func, arg1, arg2, arg3, arg4, 0); }
-static inline void xc5(smpfunc_t func, unsigned long arg1, unsigned long arg2,
-			   unsigned long arg3, unsigned long arg4, unsigned long arg5)
-{ smp_cross_call(func, arg1, arg2, arg3, arg4, arg5); }
+{ smp_cross_call(func, cpu_online_map, arg1, arg2, arg3, arg4); }
 
 static inline int smp_call_function(void (*func)(void *info), void *info, int wait)
 {
@@ -78,6 +75,14 @@
 	return 0;
 }
 
+static inline int smp_call_function_single(int cpuid, void (*func) (void *info),
+					   void *info, int wait)
+{
+	smp_cross_call((smpfunc_t)func, cpumask_of_cpu(cpuid),
+		       (unsigned long) info, 0, 0, 0);
+	return 0;
+}
+
 static inline int cpu_logical_map(int cpu)
 {
 	return cpu;
diff --git a/arch/sparc/kernel/of_device.c b/arch/sparc/kernel/of_device.c
index 4ef1607..c88af7e 100644
--- a/arch/sparc/kernel/of_device.c
+++ b/arch/sparc/kernel/of_device.c
@@ -93,7 +93,7 @@
 				       int *addrc, int *sizec);
 	int		(*map)(u32 *addr, const u32 *range,
 			       int na, int ns, int pna);
-	unsigned int	(*get_flags)(const u32 *addr);
+	unsigned long	(*get_flags)(const u32 *addr, unsigned long);
 };
 
 /*
@@ -153,8 +153,10 @@
 	return 0;
 }
 
-static unsigned int of_bus_default_get_flags(const u32 *addr)
+static unsigned long of_bus_default_get_flags(const u32 *addr, unsigned long flags)
 {
+	if (flags)
+		return flags;
 	return IORESOURCE_MEM;
 }
 
@@ -217,17 +219,21 @@
 	return 0;
 }
 
-static unsigned int of_bus_pci_get_flags(const u32 *addr)
+static unsigned long of_bus_pci_get_flags(const u32 *addr, unsigned long flags)
 {
-	unsigned int flags = 0;
 	u32 w = addr[0];
 
+	/* For PCI, we override whatever child busses may have used.  */
+	flags = 0;
 	switch((w >> 24) & 0x03) {
 	case 0x01:
 		flags |= IORESOURCE_IO;
+		break;
+
 	case 0x02: /* 32 bits */
 	case 0x03: /* 64 bits */
 		flags |= IORESOURCE_MEM;
+		break;
 	}
 	if (w & 0x40000000)
 		flags |= IORESOURCE_PREFETCH;
@@ -406,10 +412,11 @@
 		int pna, pns;
 
 		size = of_read_addr(reg + na, ns);
-		flags = bus->get_flags(reg);
 
 		memcpy(addr, reg, na * 4);
 
+		flags = bus->get_flags(reg, 0);
+
 		if (use_1to1_mapping(pp)) {
 			result = of_read_addr(addr, na);
 			goto build_res;
@@ -434,6 +441,8 @@
 					       dna, dns, pna))
 				break;
 
+			flags = pbus->get_flags(addr, flags);
+
 			dna = pna;
 			dns = pns;
 			dbus = pbus;
diff --git a/arch/sparc/kernel/sun4d_smp.c b/arch/sparc/kernel/sun4d_smp.c
index bf7147a..72fa8da 100644
--- a/arch/sparc/kernel/sun4d_smp.c
+++ b/arch/sparc/kernel/sun4d_smp.c
@@ -261,8 +261,9 @@
 static DEFINE_SPINLOCK(cross_call_lock);
 
 /* Cross calls must be serialized, at least currently. */
-void smp4d_cross_call(smpfunc_t func, unsigned long arg1, unsigned long arg2,
-		    unsigned long arg3, unsigned long arg4, unsigned long arg5)
+static void smp4d_cross_call(smpfunc_t func, cpumask_t mask, unsigned long arg1,
+			     unsigned long arg2, unsigned long arg3,
+			     unsigned long arg4)
 {
 	if(smp_processors_ready) {
 		register int high = smp_highest_cpu;
@@ -277,7 +278,7 @@
 			register unsigned long a2 asm("i2") = arg2;
 			register unsigned long a3 asm("i3") = arg3;
 			register unsigned long a4 asm("i4") = arg4;
-			register unsigned long a5 asm("i5") = arg5;
+			register unsigned long a5 asm("i5") = 0;
 
 			__asm__ __volatile__(
 				"std %0, [%6]\n\t"
@@ -289,11 +290,10 @@
 
 		/* Init receive/complete mapping, plus fire the IPI's off. */
 		{
-			cpumask_t mask;
 			register int i;
 
-			mask = cpumask_of_cpu(hard_smp4d_processor_id());
-			cpus_andnot(mask, cpu_online_map, mask);
+			cpu_clear(smp_processor_id(), mask);
+			cpus_and(mask, cpu_online_map, mask);
 			for(i = 0; i <= high; i++) {
 				if (cpu_isset(i, mask)) {
 					ccall_info.processors_in[i] = 0;
@@ -308,12 +308,16 @@
 
 			i = 0;
 			do {
+				if (!cpu_isset(i, mask))
+					continue;
 				while(!ccall_info.processors_in[i])
 					barrier();
 			} while(++i <= high);
 
 			i = 0;
 			do {
+				if (!cpu_isset(i, mask))
+					continue;
 				while(!ccall_info.processors_out[i])
 					barrier();
 			} while(++i <= high);
diff --git a/arch/sparc/kernel/sun4m_smp.c b/arch/sparc/kernel/sun4m_smp.c
index 406ac1a..a14a76a 100644
--- a/arch/sparc/kernel/sun4m_smp.c
+++ b/arch/sparc/kernel/sun4m_smp.c
@@ -244,9 +244,9 @@
 static DEFINE_SPINLOCK(cross_call_lock);
 
 /* Cross calls must be serialized, at least currently. */
-static void smp4m_cross_call(smpfunc_t func, unsigned long arg1,
+static void smp4m_cross_call(smpfunc_t func, cpumask_t mask, unsigned long arg1,
 			     unsigned long arg2, unsigned long arg3,
-			     unsigned long arg4, unsigned long arg5)
+			     unsigned long arg4)
 {
 		register int ncpus = SUN4M_NCPUS;
 		unsigned long flags;
@@ -259,14 +259,14 @@
 		ccall_info.arg2 = arg2;
 		ccall_info.arg3 = arg3;
 		ccall_info.arg4 = arg4;
-		ccall_info.arg5 = arg5;
+		ccall_info.arg5 = 0;
 
 		/* Init receive/complete mapping, plus fire the IPI's off. */
 		{
-			cpumask_t mask = cpu_online_map;
 			register int i;
 
 			cpu_clear(smp_processor_id(), mask);
+			cpus_and(mask, cpu_online_map, mask);
 			for(i = 0; i < ncpus; i++) {
 				if (cpu_isset(i, mask)) {
 					ccall_info.processors_in[i] = 0;
@@ -284,12 +284,16 @@
 
 			i = 0;
 			do {
+				if (!cpu_isset(i, mask))
+					continue;
 				while(!ccall_info.processors_in[i])
 					barrier();
 			} while(++i < ncpus);
 
 			i = 0;
 			do {
+				if (!cpu_isset(i, mask))
+					continue;
 				while(!ccall_info.processors_out[i])
 					barrier();
 			} while(++i < ncpus);
diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig
index 4a90809..5446e2a 100644
--- a/arch/sparc64/Kconfig
+++ b/arch/sparc64/Kconfig
@@ -254,6 +254,7 @@
 
 config NUMA
 	bool "NUMA support"
+	depends on SMP
 
 config NODES_SHIFT
 	int
diff --git a/arch/sparc64/kernel/of_device.c b/arch/sparc64/kernel/of_device.c
index c15bcdf..5c4fbc2 100644
--- a/arch/sparc64/kernel/of_device.c
+++ b/arch/sparc64/kernel/of_device.c
@@ -119,7 +119,7 @@
 				       int *addrc, int *sizec);
 	int		(*map)(u32 *addr, const u32 *range,
 			       int na, int ns, int pna);
-	unsigned int	(*get_flags)(const u32 *addr);
+	unsigned long	(*get_flags)(const u32 *addr, unsigned long);
 };
 
 /*
@@ -179,8 +179,10 @@
 	return 0;
 }
 
-static unsigned int of_bus_default_get_flags(const u32 *addr)
+static unsigned long of_bus_default_get_flags(const u32 *addr, unsigned long flags)
 {
+	if (flags)
+		return flags;
 	return IORESOURCE_MEM;
 }
 
@@ -272,17 +274,21 @@
 	return 0;
 }
 
-static unsigned int of_bus_pci_get_flags(const u32 *addr)
+static unsigned long of_bus_pci_get_flags(const u32 *addr, unsigned long flags)
 {
-	unsigned int flags = 0;
 	u32 w = addr[0];
 
+	/* For PCI, we override whatever child busses may have used.  */
+	flags = 0;
 	switch((w >> 24) & 0x03) {
 	case 0x01:
 		flags |= IORESOURCE_IO;
+		break;
+
 	case 0x02: /* 32 bits */
 	case 0x03: /* 64 bits */
 		flags |= IORESOURCE_MEM;
+		break;
 	}
 	if (w & 0x40000000)
 		flags |= IORESOURCE_PREFETCH;
@@ -510,10 +516,10 @@
 		int pna, pns;
 
 		size = of_read_addr(reg + na, ns);
-		flags = bus->get_flags(reg);
-
 		memcpy(addr, reg, na * 4);
 
+		flags = bus->get_flags(addr, 0);
+
 		if (use_1to1_mapping(pp)) {
 			result = of_read_addr(addr, na);
 			goto build_res;
@@ -538,6 +544,8 @@
 					       dna, dns, pna))
 				break;
 
+			flags = pbus->get_flags(addr, flags);
+
 			dna = pna;
 			dns = pns;
 			dbus = pbus;
diff --git a/arch/sparc64/kernel/smp.c b/arch/sparc64/kernel/smp.c
index 743ccad..2be166c 100644
--- a/arch/sparc64/kernel/smp.c
+++ b/arch/sparc64/kernel/smp.c
@@ -80,8 +80,6 @@
 			   i, cpu_data(i).clock_tick);
 }
 
-static __cacheline_aligned_in_smp DEFINE_SPINLOCK(call_lock);
-
 extern void setup_sparc64_timer(void);
 
 static volatile unsigned long callin_flag = 0;
@@ -120,9 +118,9 @@
 	while (!cpu_isset(cpuid, smp_commenced_mask))
 		rmb();
 
-	spin_lock(&call_lock);
+	ipi_call_lock();
 	cpu_set(cpuid, cpu_online_map);
-	spin_unlock(&call_lock);
+	ipi_call_unlock();
 
 	/* idle thread is expected to have preempt disabled */
 	preempt_disable();
@@ -1305,10 +1303,6 @@
 	c->core_id = 0;
 	c->proc_id = -1;
 
-	spin_lock(&call_lock);
-	cpu_clear(cpu, cpu_online_map);
-	spin_unlock(&call_lock);
-
 	smp_wmb();
 
 	/* Make sure no interrupts point to this cpu.  */
@@ -1318,6 +1312,10 @@
 	mdelay(1);
 	local_irq_disable();
 
+	ipi_call_lock();
+	cpu_clear(cpu, cpu_online_map);
+	ipi_call_unlock();
+
 	return 0;
 }
 
diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c
index 7c3c81f..4fb9de0 100644
--- a/arch/sparc64/mm/init.c
+++ b/arch/sparc64/mm/init.c
@@ -1875,7 +1875,7 @@
  * memory list again, and make sure it provides at least as much
  * memory as 'pavail' does.
  */
-static void setup_valid_addr_bitmap_from_pavail(void)
+static void __init setup_valid_addr_bitmap_from_pavail(void)
 {
 	int i;