Merge master.kernel.org:/home/rmk/linux-2.6-arm

16 files changed, 358 insertions, 54 deletions

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 296bc03d1cf1..91d5ef3397be 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -324,7 +324,7 @@ menu "Kernel Features"
 
 config SMP
         bool "Symmetric Multi-Processing (EXPERIMENTAL)"
-        depends on EXPERIMENTAL && BROKEN #&& n
+        depends on EXPERIMENTAL && REALVIEW_MPCORE
         help
           This enables support for systems with more than one CPU. If you have
           a system with only one CPU, like most personal computers, say N. If
@@ -585,7 +585,7 @@ config FPE_NWFPE
 
 config FPE_NWFPE_XP
         bool "Support extended precision"
-        depends on FPE_NWFPE && !CPU_BIG_ENDIAN
+        depends on FPE_NWFPE
         help
           Say Y to include 80-bit support in the kernel floating-point
           emulator.  Otherwise, only 32 and 64-bit support is compiled in.
diff --git a/arch/arm/mach-ixp2000/core.c b/arch/arm/mach-ixp2000/core.c
index df140962bb0f..6851abaf5524 100644
--- a/arch/arm/mach-ixp2000/core.c
+++ b/arch/arm/mach-ixp2000/core.c
@@ -84,63 +84,54 @@ static struct map_desc ixp2000_io_desc[] __initdata = {
                 .virtual        = IXP2000_CAP_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_CAP_PHYS_BASE),
                 .length         = IXP2000_CAP_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_INTCTL_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_INTCTL_PHYS_BASE),
                 .length         = IXP2000_INTCTL_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_PCI_CREG_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_PCI_CREG_PHYS_BASE),
                 .length         = IXP2000_PCI_CREG_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_PCI_CSR_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_PCI_CSR_PHYS_BASE),
                 .length         = IXP2000_PCI_CSR_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_MSF_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_MSF_PHYS_BASE),
                 .length         = IXP2000_MSF_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_PCI_IO_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_PCI_IO_PHYS_BASE),
                 .length         = IXP2000_PCI_IO_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_PCI_CFG0_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_PCI_CFG0_PHYS_BASE),
                 .length         = IXP2000_PCI_CFG0_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }, {
                 .virtual        = IXP2000_PCI_CFG1_VIRT_BASE,
                 .pfn            = __phys_to_pfn(IXP2000_PCI_CFG1_PHYS_BASE),
                 .length         = IXP2000_PCI_CFG1_SIZE,
-                .type           = MT_DEVICE
+                .type           = MT_IXP2000_DEVICE,
         }
 };
 
 void __init ixp2000_map_io(void)
 {
-        extern unsigned int processor_id;
-
         /*
-         * On IXP2400 CPUs we need to use MT_IXP2000_DEVICE for
-         * tweaking the PMDs so XCB=101. On IXP2800s we use the normal
-         * PMD flags.
+         * On IXP2400 CPUs we need to use MT_IXP2000_DEVICE so that
+         * XCB=101 (to avoid triggering erratum #66), and given that
+         * this mode speeds up I/O accesses and we have write buffer
+         * flushes in the right places anyway, it doesn't hurt to use
+         * XCB=101 for all IXP2000s.
          */
-        if ((processor_id & 0xfffffff0) == 0x69054190) {
-                int i;
-
-                printk(KERN_INFO "Enabling IXP2400 erratum #66 workaround\n");
-
-                for(i=0;i<ARRAY_SIZE(ixp2000_io_desc);i++)
-                        ixp2000_io_desc[i].type = MT_IXP2000_DEVICE;
-        }
-
         iotable_init(ixp2000_io_desc, ARRAY_SIZE(ixp2000_io_desc));
 
         /* Set slowport to 8-bit mode.  */
diff --git a/arch/arm/mach-realview/Kconfig b/arch/arm/mach-realview/Kconfig
index 4b63dc9eabfe..129976866d47 100644
--- a/arch/arm/mach-realview/Kconfig
+++ b/arch/arm/mach-realview/Kconfig
@@ -8,4 +8,13 @@ config MACH_REALVIEW_EB
         help
           Include support for the ARM(R) RealView Emulation Baseboard platform.
 
+config REALVIEW_MPCORE
+        bool "Support MPcore tile"
+        depends on MACH_REALVIEW_EB
+        help
+          Enable support for the MPCore tile on the Realview platform.
+          Since there are device address and interrupt differences, a
+          kernel built with this option enabled is not compatible with
+          other tiles.
+
 endmenu
diff --git a/arch/arm/mach-realview/Makefile b/arch/arm/mach-realview/Makefile
index 8d37ea1605fd..011a85c10627 100644
--- a/arch/arm/mach-realview/Makefile
+++ b/arch/arm/mach-realview/Makefile
@@ -4,3 +4,4 @@
 
 obj-y                                   := core.o clock.o
 obj-$(CONFIG_MACH_REALVIEW_EB)          += realview_eb.o
+obj-$(CONFIG_SMP)                       += platsmp.o headsmp.o
diff --git a/arch/arm/mach-realview/core.h b/arch/arm/mach-realview/core.h
index 575599db74db..d83e8bad2038 100644
--- a/arch/arm/mach-realview/core.h
+++ b/arch/arm/mach-realview/core.h
@@ -23,6 +23,7 @@
 #define __ASM_ARCH_REALVIEW_H
 
 #include <asm/hardware/amba.h>
+#include <asm/leds.h>
 #include <asm/io.h>
 
 #define __io_address(n)         __io(IO_ADDRESS(n))
diff --git a/arch/arm/mach-realview/headsmp.S b/arch/arm/mach-realview/headsmp.S
new file mode 100644
index 000000000000..4075473cf68a
--- /dev/null
+++ b/arch/arm/mach-realview/headsmp.S
@@ -0,0 +1,39 @@
+/*
+ *  linux/arch/arm/mach-realview/headsmp.S
+ *
+ *  Copyright (c) 2003 ARM Limited
+ *  All Rights Reserved
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/linkage.h>
+#include <linux/init.h>
+
+        __INIT
+
+/*
+ * Realview specific entry point for secondary CPUs.  This provides
+ * a "holding pen" into which all secondary cores are held until we're
+ * ready for them to initialise.
+ */
+ENTRY(realview_secondary_startup)
+        mrc     p15, 0, r0, c0, c0, 5
+        and     r0, r0, #15
+        adr     r4, 1f
+        ldmia   r4, {r5, r6}
+        sub     r4, r4, r5
+        add     r6, r6, r4
+pen:    ldr     r7, [r6]
+        cmp     r7, r0
+        bne     pen
+
+        /*
+         * we've been released from the holding pen: secondary_stack
+         * should now contain the SVC stack for this core
+         */
+        b       secondary_startup
+
+1:      .long   .
+        .long   pen_release
diff --git a/arch/arm/mach-realview/platsmp.c b/arch/arm/mach-realview/platsmp.c
new file mode 100644
index 000000000000..9844644d0fb5
--- /dev/null
+++ b/arch/arm/mach-realview/platsmp.c
@@ -0,0 +1,195 @@
+/*
+ *  linux/arch/arm/mach-realview/platsmp.c
+ *
+ *  Copyright (C) 2002 ARM Ltd.
+ *  All Rights Reserved
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/smp.h>
+
+#include <asm/cacheflush.h>
+#include <asm/hardware/arm_scu.h>
+#include <asm/hardware.h>
+
+#include "core.h"
+
+extern void realview_secondary_startup(void);
+
+/*
+ * control for which core is the next to come out of the secondary
+ * boot "holding pen"
+ */
+volatile int __cpuinitdata pen_release = -1;
+
+static unsigned int __init get_core_count(void)
+{
+        unsigned int ncores;
+
+        ncores = __raw_readl(IO_ADDRESS(REALVIEW_MPCORE_SCU_BASE) + SCU_CONFIG);
+
+        return (ncores & 0x03) + 1;
+}
+
+static DEFINE_SPINLOCK(boot_lock);
+
+void __cpuinit platform_secondary_init(unsigned int cpu)
+{
+        /*
+         * the primary core may have used a "cross call" soft interrupt
+         * to get this processor out of WFI in the BootMonitor - make
+         * sure that we are no longer being sent this soft interrupt
+         */
+        smp_cross_call_done(cpumask_of_cpu(cpu));
+
+        /*
+         * if any interrupts are already enabled for the primary
+         * core (e.g. timer irq), then they will not have been enabled
+         * for us: do so
+         */
+        gic_cpu_init(__io_address(REALVIEW_GIC_CPU_BASE));
+
+        /*
+         * let the primary processor know we're out of the
+         * pen, then head off into the C entry point
+         */
+        pen_release = -1;
+
+        /*
+         * Synchronise with the boot thread.
+         */
+        spin_lock(&boot_lock);
+        spin_unlock(&boot_lock);
+}
+
+int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+{
+        unsigned long timeout;
+
+        /*
+         * set synchronisation state between this boot processor
+         * and the secondary one
+         */
+        spin_lock(&boot_lock);
+
+        /*
+         * The secondary processor is waiting to be released from
+         * the holding pen - release it, then wait for it to flag
+         * that it has been released by resetting pen_release.
+         *
+         * Note that "pen_release" is the hardware CPU ID, whereas
+         * "cpu" is Linux's internal ID.
+         */
+        pen_release = cpu;
+        flush_cache_all();
+
+        /*
+         * XXX
+         *
+         * This is a later addition to the booting protocol: the
+         * bootMonitor now puts secondary cores into WFI, so
+         * poke_milo() no longer gets the cores moving; we need
+         * to send a soft interrupt to wake the secondary core.
+         * Use smp_cross_call() for this, since there's little
+         * point duplicating the code here
+         */
+        smp_cross_call(cpumask_of_cpu(cpu));
+
+        timeout = jiffies + (1 * HZ);
+        while (time_before(jiffies, timeout)) {
+                if (pen_release == -1)
+                        break;
+
+                udelay(10);
+        }
+
+        /*
+         * now the secondary core is starting up let it run its
+         * calibrations, then wait for it to finish
+         */
+        spin_unlock(&boot_lock);
+
+        return pen_release != -1 ? -ENOSYS : 0;
+}
+
+static void __init poke_milo(void)
+{
+        extern void secondary_startup(void);
+
+        /* nobody is to be released from the pen yet */
+        pen_release = -1;
+
+        /*
+         * write the address of secondary startup into the system-wide
+         * flags register, then clear the bottom two bits, which is what
+         * BootMonitor is waiting for
+         */
+#if 1
+#define REALVIEW_SYS_FLAGSS_OFFSET 0x30
+        __raw_writel(virt_to_phys(realview_secondary_startup),
+                     (IO_ADDRESS(REALVIEW_SYS_BASE) +
+                      REALVIEW_SYS_FLAGSS_OFFSET));
+#define REALVIEW_SYS_FLAGSC_OFFSET 0x34
+        __raw_writel(3,
+                     (IO_ADDRESS(REALVIEW_SYS_BASE) +
+                      REALVIEW_SYS_FLAGSC_OFFSET));
+#endif
+
+        mb();
+}
+
+void __init smp_prepare_cpus(unsigned int max_cpus)
+{
+        unsigned int ncores = get_core_count();
+        unsigned int cpu = smp_processor_id();
+        int i;
+
+        /* sanity check */
+        if (ncores == 0) {
+                printk(KERN_ERR
+                       "Realview: strange CM count of 0? Default to 1\n");
+
+                ncores = 1;
+        }
+
+        if (ncores > NR_CPUS) {
+                printk(KERN_WARNING
+                       "Realview: no. of cores (%d) greater than configured "
+                       "maximum of %d - clipping\n",
+                       ncores, NR_CPUS);
+                ncores = NR_CPUS;
+        }
+
+        smp_store_cpu_info(cpu);
+
+        /*
+         * are we trying to boot more cores than exist?
+         */
+        if (max_cpus > ncores)
+                max_cpus = ncores;
+
+        /*
+         * Initialise the possible/present maps.
+         * cpu_possible_map describes the set of CPUs which may be present
+         * cpu_present_map describes the set of CPUs populated
+         */
+        for (i = 0; i < max_cpus; i++) {
+                cpu_set(i, cpu_possible_map);
+                cpu_set(i, cpu_present_map);
+        }
+
+        /*
+         * Do we need any more CPUs? If so, then let them know where
+         * to start. Note that, on modern versions of MILO, the "poke"
+         * doesn't actually do anything until each individual core is
+         * sent a soft interrupt to get it out of WFI
+         */
+        if (max_cpus > 1)
+                poke_milo();
+}
diff --git a/arch/arm/mach-realview/realview_eb.c b/arch/arm/mach-realview/realview_eb.c
index 267bb07e39b7..7dc32503fdf2 100644
--- a/arch/arm/mach-realview/realview_eb.c
+++ b/arch/arm/mach-realview/realview_eb.c
@@ -136,6 +136,11 @@ static struct amba_device *amba_devs[] __initdata = {
 
 static void __init gic_init_irq(void)
 {
+#ifdef CONFIG_REALVIEW_MPCORE
+        writel(0x0000a05f, __io_address(REALVIEW_SYS_LOCK));
+        writel(0x008003c0, __io_address(REALVIEW_SYS_BASE) + 0xd8);
+        writel(0x00000000, __io_address(REALVIEW_SYS_LOCK));
+#endif
         gic_dist_init(__io_address(REALVIEW_GIC_DIST_BASE));
         gic_cpu_init(__io_address(REALVIEW_GIC_CPU_BASE));
 }
diff --git a/arch/arm/mm/mm-armv.c b/arch/arm/mm/mm-armv.c
index fb5b40289de2..9e50127be635 100644
--- a/arch/arm/mm/mm-armv.c
+++ b/arch/arm/mm/mm-armv.c
@@ -354,7 +354,7 @@ void __init build_mem_type_table(void)
 {
         struct cachepolicy *cp;
         unsigned int cr = get_cr();
-        unsigned int user_pgprot;
+        unsigned int user_pgprot, kern_pgprot;
         int cpu_arch = cpu_architecture();
         int i;
 
@@ -381,7 +381,7 @@ void __init build_mem_type_table(void)
         }
 
         cp = &cache_policies[cachepolicy];
-        user_pgprot = cp->pte;
+        kern_pgprot = user_pgprot = cp->pte;
 
         /*
          * ARMv6 and above have extended page tables.
@@ -393,6 +393,7 @@ void __init build_mem_type_table(void)
                  */
                 mem_types[MT_MEMORY].prot_sect &= ~PMD_BIT4;
                 mem_types[MT_ROM].prot_sect &= ~PMD_BIT4;
+
                 /*
                  * Mark cache clean areas and XIP ROM read only
                  * from SVC mode and no access from userspace.
@@ -412,32 +413,47 @@ void __init build_mem_type_table(void)
                  * (iow, non-global)
                  */
                 user_pgprot |= L_PTE_ASID;
+
+#ifdef CONFIG_SMP
+                /*
+                 * Mark memory with the "shared" attribute for SMP systems
+                 */
+                user_pgprot |= L_PTE_SHARED;
+                kern_pgprot |= L_PTE_SHARED;
+                mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S;
+#endif
         }
 
+        for (i = 0; i < 16; i++) {
+                unsigned long v = pgprot_val(protection_map[i]);
+                v = (v & ~(L_PTE_BUFFERABLE|L_PTE_CACHEABLE)) | user_pgprot;
+                protection_map[i] = __pgprot(v);
+        }
+
+        mem_types[MT_LOW_VECTORS].prot_pte |= kern_pgprot;
+        mem_types[MT_HIGH_VECTORS].prot_pte |= kern_pgprot;
+
         if (cpu_arch >= CPU_ARCH_ARMv5) {
-                mem_types[MT_LOW_VECTORS].prot_pte |= cp->pte & PTE_CACHEABLE;
-                mem_types[MT_HIGH_VECTORS].prot_pte |= cp->pte & PTE_CACHEABLE;
+#ifndef CONFIG_SMP
+                /*
+                 * Only use write-through for non-SMP systems
+                 */
+                mem_types[MT_LOW_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE;
+                mem_types[MT_HIGH_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE;
+#endif
         } else {
-                mem_types[MT_LOW_VECTORS].prot_pte |= cp->pte;
-                mem_types[MT_HIGH_VECTORS].prot_pte |= cp->pte;
                 mem_types[MT_MINICLEAN].prot_sect &= ~PMD_SECT_TEX(1);
         }
 
+        pgprot_kernel = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG |
+                                 L_PTE_DIRTY | L_PTE_WRITE |
+                                 L_PTE_EXEC | kern_pgprot);
+
         mem_types[MT_LOW_VECTORS].prot_l1 |= ecc_mask;
         mem_types[MT_HIGH_VECTORS].prot_l1 |= ecc_mask;
         mem_types[MT_MEMORY].prot_sect |= ecc_mask | cp->pmd;
         mem_types[MT_ROM].prot_sect |= cp->pmd;
 
-        for (i = 0; i < 16; i++) {
-                unsigned long v = pgprot_val(protection_map[i]);
-                v = (v & ~(PTE_BUFFERABLE|PTE_CACHEABLE)) | user_pgprot;
-                protection_map[i] = __pgprot(v);
-        }
-
-        pgprot_kernel = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG |
-                                 L_PTE_DIRTY | L_PTE_WRITE |
-                                 L_PTE_EXEC | cp->pte);
-
         switch (cp->pmd) {
         case PMD_SECT_WT:
                 mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WT;
diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S
index 9bb5fff406fb..92f3ca31b7b9 100644
--- a/arch/arm/mm/proc-v6.S
+++ b/arch/arm/mm/proc-v6.S
@@ -12,6 +12,7 @@
 #include <linux/linkage.h>
 #include <asm/assembler.h>
 #include <asm/asm-offsets.h>
+#include <asm/hardware/arm_scu.h>
 #include <asm/procinfo.h>
 #include <asm/pgtable.h>
 
@@ -112,6 +113,9 @@ ENTRY(cpu_v6_dcache_clean_area)
 ENTRY(cpu_v6_switch_mm)
         mov     r2, #0
         ldr     r1, [r1, #MM_CONTEXT_ID]        @ get mm->context.id
+#ifdef CONFIG_SMP
+        orr     r0, r0, #2                      @ set shared pgtable
+#endif
         mcr     p15, 0, r2, c7, c5, 6           @ flush BTAC/BTB
         mcr     p15, 0, r2, c7, c10, 4          @ drain write buffer
         mcr     p15, 0, r0, c2, c0, 0           @ set TTB 0
@@ -140,7 +144,7 @@ ENTRY(cpu_v6_switch_mm)
 ENTRY(cpu_v6_set_pte)
         str     r1, [r0], #-2048                @ linux version
 
-        bic     r2, r1, #0x000007f0
+        bic     r2, r1, #0x000003f0
         bic     r2, r2, #0x00000003
         orr     r2, r2, #PTE_EXT_AP0 | 2
 
@@ -191,6 +195,23 @@ cpu_v6_name:
  *      - cache type register is implemented
  */
 __v6_setup:
+#ifdef CONFIG_SMP
+        /* Set up the SCU on core 0 only */
+        mrc     p15, 0, r0, c0, c0, 5           @ CPU core number
+        ands    r0, r0, #15
+        moveq   r0, #0x10000000 @ SCU_BASE
+        orreq   r0, r0, #0x00100000
+        ldreq   r5, [r0, #SCU_CTRL]
+        orreq   r5, r5, #1
+        streq   r5, [r0, #SCU_CTRL]
+
+#ifndef CONFIG_CPU_DCACHE_DISABLE
+        mrc     p15, 0, r0, c1, c0, 1           @ Enable SMP/nAMP mode
+        orr     r0, r0, #0x20
+        mcr     p15, 0, r0, c1, c0, 1
+#endif
+#endif
+
         mov     r0, #0
         mcr     p15, 0, r0, c7, c14, 0          @ clean+invalidate D cache
         mcr     p15, 0, r0, c7, c5, 0           @ invalidate I cache
@@ -198,6 +219,9 @@ __v6_setup:
         mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
         mcr     p15, 0, r0, c8, c7, 0           @ invalidate I + D TLBs
         mcr     p15, 0, r0, c2, c0, 2           @ TTB control register
+#ifdef CONFIG_SMP
+        orr     r4, r4, #2                      @ set shared pgtable
+#endif
         mcr     p15, 0, r4, c2, c0, 1           @ load TTB1
 #ifdef CONFIG_VFP
         mrc     p15, 0, r0, c1, c0, 2
diff --git a/arch/arm/nwfpe/fpa11.h b/arch/arm/nwfpe/fpa11.h
index 9677ae8448e8..da4c616b6c49 100644
--- a/arch/arm/nwfpe/fpa11.h
+++ b/arch/arm/nwfpe/fpa11.h
@@ -60,7 +60,7 @@ typedef union tagFPREG {
 #ifdef CONFIG_FPE_NWFPE_XP
         floatx80 fExtended;
 #else
-        int padding[3];
+        u32 padding[3];
 #endif
 } FPREG;
 
diff --git a/arch/arm/nwfpe/fpa11_cpdt.c b/arch/arm/nwfpe/fpa11_cpdt.c
index b0db5cbcc3b1..32859fa8dcfc 100644
--- a/arch/arm/nwfpe/fpa11_cpdt.c
+++ b/arch/arm/nwfpe/fpa11_cpdt.c
@@ -59,8 +59,13 @@ static inline void loadExtended(const unsigned int Fn, const unsigned int __user
         p = (unsigned int *) &fpa11->fpreg[Fn].fExtended;
         fpa11->fType[Fn] = typeExtended;
         get_user(p[0], &pMem[0]);       /* sign & exponent */
+#ifdef __ARMEB__
+        get_user(p[1], &pMem[1]);       /* ms bits */
+        get_user(p[2], &pMem[2]);       /* ls bits */
+#else
         get_user(p[1], &pMem[2]);       /* ls bits */
         get_user(p[2], &pMem[1]);       /* ms bits */
+#endif
 }
 #endif
 
@@ -177,8 +182,13 @@ static inline void storeExtended(const unsigned int Fn, unsigned int __user *pMe
         }
 
         put_user(val.i[0], &pMem[0]);   /* sign & exp */
+#ifdef __ARMEB__
+        put_user(val.i[1], &pMem[1]);   /* msw */
+        put_user(val.i[2], &pMem[2]);
+#else
         put_user(val.i[1], &pMem[2]);
         put_user(val.i[2], &pMem[1]);   /* msw */
+#endif
 }
 #endif
 
diff --git a/arch/arm/nwfpe/fpopcode.c b/arch/arm/nwfpe/fpopcode.c
index 4c9f5703148c..67ff2ab08ea0 100644
--- a/arch/arm/nwfpe/fpopcode.c
+++ b/arch/arm/nwfpe/fpopcode.c
@@ -29,14 +29,14 @@
 
 #ifdef CONFIG_FPE_NWFPE_XP
 const floatx80 floatx80Constant[] = {
-        {0x0000, 0x0000000000000000ULL},        /* extended 0.0 */
-        {0x3fff, 0x8000000000000000ULL},        /* extended 1.0 */
-        {0x4000, 0x8000000000000000ULL},        /* extended 2.0 */
-        {0x4000, 0xc000000000000000ULL},        /* extended 3.0 */
-        {0x4001, 0x8000000000000000ULL},        /* extended 4.0 */
-        {0x4001, 0xa000000000000000ULL},        /* extended 5.0 */
-        {0x3ffe, 0x8000000000000000ULL},        /* extended 0.5 */
-        {0x4002, 0xa000000000000000ULL}         /* extended 10.0 */
+        { .high = 0x0000, .low = 0x0000000000000000ULL},/* extended 0.0 */
+        { .high = 0x3fff, .low = 0x8000000000000000ULL},/* extended 1.0 */
+        { .high = 0x4000, .low = 0x8000000000000000ULL},/* extended 2.0 */
+        { .high = 0x4000, .low = 0xc000000000000000ULL},/* extended 3.0 */
+        { .high = 0x4001, .low = 0x8000000000000000ULL},/* extended 4.0 */
+        { .high = 0x4001, .low = 0xa000000000000000ULL},/* extended 5.0 */
+        { .high = 0x3ffe, .low = 0x8000000000000000ULL},/* extended 0.5 */
+        { .high = 0x4002, .low = 0xa000000000000000ULL},/* extended 10.0 */
 };
 #endif
 
diff --git a/arch/arm/nwfpe/softfloat-specialize b/arch/arm/nwfpe/softfloat-specialize
index acf409144763..d4a4c8e06635 100644
--- a/arch/arm/nwfpe/softfloat-specialize
+++ b/arch/arm/nwfpe/softfloat-specialize
@@ -332,6 +332,7 @@ static floatx80 commonNaNToFloatx80( commonNaNT a )
 
     z.low = LIT64( 0xC000000000000000 ) | ( a.high>>1 );
     z.high = ( ( (bits16) a.sign )<<15 ) | 0x7FFF;
+    z.__padding = 0;
     return z;
 
 }
diff --git a/arch/arm/nwfpe/softfloat.c b/arch/arm/nwfpe/softfloat.c
index f9f049132a17..0f9656e482ba 100644
--- a/arch/arm/nwfpe/softfloat.c
+++ b/arch/arm/nwfpe/softfloat.c
@@ -531,6 +531,7 @@ INLINE floatx80 packFloatx80( flag zSign, int32 zExp, bits64 zSig )
 
     z.low = zSig;
     z.high = ( ( (bits16) zSign )<<15 ) + zExp;
+    z.__padding = 0;
     return z;
 
 }
@@ -2831,6 +2832,7 @@ static floatx80 subFloatx80Sigs( struct roundingData *roundData, floatx80 a, flo
         roundData->exception |= float_flag_invalid;
         z.low = floatx80_default_nan_low;
         z.high = floatx80_default_nan_high;
+        z.__padding = 0;
         return z;
     }
     if ( aExp == 0 ) {
@@ -2950,6 +2952,7 @@ floatx80 floatx80_mul( struct roundingData *roundData, floatx80 a, floatx80 b )
             roundData->exception |= float_flag_invalid;
             z.low = floatx80_default_nan_low;
             z.high = floatx80_default_nan_high;
+            z.__padding = 0;
             return z;
         }
         return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) );
@@ -3015,6 +3018,7 @@ floatx80 floatx80_div( struct roundingData *roundData, floatx80 a, floatx80 b )
                 roundData->exception |= float_flag_invalid;
                 z.low = floatx80_default_nan_low;
                 z.high = floatx80_default_nan_high;
+                z.__padding = 0;
                 return z;
             }
             roundData->exception |= float_flag_divbyzero;
@@ -3093,6 +3097,7 @@ floatx80 floatx80_rem( struct roundingData *roundData, floatx80 a, floatx80 b )
             roundData->exception |= float_flag_invalid;
             z.low = floatx80_default_nan_low;
             z.high = floatx80_default_nan_high;
+            z.__padding = 0;
             return z;
         }
         normalizeFloatx80Subnormal( bSig, &bExp, &bSig );
@@ -3184,6 +3189,7 @@ floatx80 floatx80_sqrt( struct roundingData *roundData, floatx80 a )
         roundData->exception |= float_flag_invalid;
         z.low = floatx80_default_nan_low;
         z.high = floatx80_default_nan_high;
+        z.__padding = 0;
         return z;
     }
     if ( aExp == 0 ) {
diff --git a/arch/arm/nwfpe/softfloat.h b/arch/arm/nwfpe/softfloat.h
index 14151700b6b2..978c699673c6 100644
--- a/arch/arm/nwfpe/softfloat.h
+++ b/arch/arm/nwfpe/softfloat.h
@@ -51,11 +51,17 @@ input or output the `floatx80' type will be defined.
 Software IEC/IEEE floating-point types.
 -------------------------------------------------------------------------------
 */
-typedef unsigned long int float32;
-typedef unsigned long long float64;
+typedef u32 float32;
+typedef u64 float64;
 typedef struct {
-    unsigned short high;
-    unsigned long long low;
+#ifdef __ARMEB__
+    u16 __padding;
+    u16 high;
+#else
+    u16 high;
+    u16 __padding;
+#endif
+    u64 low;
 } floatx80;
 
 /*